)]}' { "log": [ { "commit": "84196390620ac0e5070ae36af84c137c6216a7dc", "tree": "26d3132b87fc8510c3f48872c7be5e7b3651904d", "parents": [ "0d02ec6b3136c73c09e7859f0d0e4e2c4c07b49b", "ee5de60a08b7d8d255722662da461ea159c15538" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Mar 22 11:34:31 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Mar 22 11:34:31 2021 -0700" }, "message": "Merge tag \u0027selinux-pr-20210322\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/selinux\n\nPull selinux fixes from Paul Moore:\n \"Three SELinux patches:\n\n - Fix a problem where a local variable is used outside its associated\n function. Thankfully this can only be triggered by reloading the\n SELinux policy, which is a restricted operation for other obvious\n reasons.\n\n - Fix some incorrect, and inconsistent, audit and printk messages\n when loading the SELinux policy.\n\n All three patches are relatively minor and have been through our\n testing with no failures\"\n\n* tag \u0027selinux-pr-20210322\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/selinux:\n selinuxfs: unify policy load error reporting\n selinux: fix variable scope issue in live sidtab conversion\n selinux: don\u0027t log MAC_POLICY_LOAD record on failed policy load\n" }, { "commit": "0d02ec6b3136c73c09e7859f0d0e4e2c4c07b49b", "tree": "76908f05975a5d8019b8bc10836fd69f20d17a30", "parents": [ "d7f5f1bd3c240c4d527c0871a38dc3d61255ea9e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 14:56:43 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 14:56:43 2021 -0700" }, "message": "Linux 5.12-rc4\n" }, { "commit": "d7f5f1bd3c240c4d527c0871a38dc3d61255ea9e", "tree": "7ed2121771fe6c80854edda6bba7daa3d625b38f", "parents": [ "2c41fab1c60b02626c8153a1806a7a1e5d62aaf1", "64395d950bc476106b39341e42ebfd4d2eb71d2c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 14:06:10 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 14:06:10 2021 -0700" }, "message": "Merge tag \u0027ext4_for_linus_stable\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4\n\nPull ext4 fixes from Ted Ts\u0027o:\n \"Miscellaneous ext4 bug fixes for v5.12\"\n\n* tag \u0027ext4_for_linus_stable\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:\n ext4: initialize ret to suppress smatch warning\n ext4: stop inode update before return\n ext4: fix rename whiteout with fast commit\n ext4: fix timer use-after-free on failed mount\n ext4: fix potential error in ext4_do_update_inode\n ext4: do not try to set xattr into ea_inode if value is empty\n ext4: do not iput inode under running transaction in ext4_rename()\n ext4: find old entry again if failed to rename whiteout\n ext4: fix error handling in ext4_end_enable_verity()\n ext4: fix bh ref count on error paths\n fs/ext4: fix integer overflow in s_log_groups_per_flex\n ext4: add reclaim checks to xattr code\n ext4: shrink race window in ext4_should_retry_alloc()\n" }, { "commit": "2c41fab1c60b02626c8153a1806a7a1e5d62aaf1", "tree": "f13c693ef22d4fd7b181846c0a45f0a4cea9943e", "parents": [ "1d4345eb51a185b2ecd11834dbddca79cb922eb5", "0031275d119efe16711cd93519b595e6f9b4b330" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 12:25:54 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 12:25:54 2021 -0700" }, "message": "Merge tag \u0027io_uring-5.12-2021-03-21\u0027 of git://git.kernel.dk/linux-block\n\nPull io_uring followup fixes from Jens Axboe:\n\n - The SIGSTOP change from Eric, so we properly ignore that for\n PF_IO_WORKER threads.\n\n - Disallow sending signals to PF_IO_WORKER threads in general, we\u0027re\n not interested in having them funnel back to the io_uring owning\n task.\n\n - Stable fix from Stefan, ensuring we properly break links for short\n send/sendmsg recv/recvmsg if MSG_WAITALL is set.\n\n - Catch and loop when needing to run task_work before a PF_IO_WORKER\n threads goes to sleep.\n\n* tag \u0027io_uring-5.12-2021-03-21\u0027 of git://git.kernel.dk/linux-block:\n io_uring: call req_set_fail_links() on short send[msg]()/recv[msg]() with MSG_WAITALL\n io-wq: ensure task is running before processing task_work\n signal: don\u0027t allow STOP on PF_IO_WORKER threads\n signal: don\u0027t allow sending any signals to PF_IO_WORKER threads\n" }, { "commit": "1d4345eb51a185b2ecd11834dbddca79cb922eb5", "tree": "496b893e3d3714def5a79647ee6ff68c4893d916", "parents": [ "3001c3554f1dfac3f2561a6b34c5d209e3592dc0", "2cafd46a714af1e55354bc6dcea9dcc13f9475b5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:54:04 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:54:04 2021 -0700" }, "message": "Merge tag \u0027staging-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging\n\nPull staging and IIO driver fixes from Greg KH:\n \"Some small staging and IIO driver fixes:\n\n - MAINTAINERS changes for the move of the staging mailing list\n\n - comedi driver fixes to get request_irq() to work correctly\n\n - counter driver fixes for reported issues with iio devices\n\n - tiny iio driver fixes for reported issues.\n\n All of these have been in linux-next with no reported problems\"\n\n* tag \u0027staging-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging:\n staging: vt665x: fix alignment constraints\n staging: comedi: cb_pcidas64: fix request_irq() warn\n staging: comedi: cb_pcidas: fix request_irq() warn\n MAINTAINERS: move the staging subsystem to lists.linux.dev\n MAINTAINERS: move some real subsystems off of the staging mailing list\n iio: gyro: mpu3050: Fix error handling in mpu3050_trigger_handler\n iio: hid-sensor-temperature: Fix issues of timestamp channel\n iio: hid-sensor-humidity: Fix alignment issue of timestamp channel\n counter: stm32-timer-cnt: fix ceiling miss-alignment with reload register\n counter: stm32-timer-cnt: fix ceiling write max value\n counter: stm32-timer-cnt: Report count function when SLAVE_MODE_DISABLED\n iio: adc: ab8500-gpadc: Fix off by 10 to 3\n iio:adc:stm32-adc: Add HAS_IOMEM dependency\n iio: adis16400: Fix an error code in adis16400_initial_setup()\n iio: adc: adi-axi-adc: add proper Kconfig dependencies\n iio: adc: ad7949: fix wrong ADC result due to incorrect bit mask\n iio: hid-sensor-prox: Fix scale not correct issue\n iio:adc:qcom-spmi-vadc: add default scale to LR_MUX2_BAT_ID channel\n" }, { "commit": "3001c3554f1dfac3f2561a6b34c5d209e3592dc0", "tree": "25e92a53a3bb82d4abbb0bc8313820de98b975a4", "parents": [ "5ee96fa9dd78b2dc81b587e33074d877002d7605", "2b8c956ea6ba896ec18ae36c2684ecfa04c1f479" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:49:16 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:49:16 2021 -0700" }, "message": "Merge tag \u0027usb-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb\n\nPull USB and Thunderbolt driver fixes from Greg KH:\n \"Here are some small Thunderbolt and USB driver fixes for some reported\n issues:\n\n - thunderbolt fixes for minor problems\n\n - typec fixes for power issues\n\n - usb-storage quirk addition\n\n - usbip bugfix\n\n - dwc3 bugfix when stopping transfers\n\n - cdnsp bugfix for isoc transfers\n\n - gadget use-after-free fix\n\n All have been in linux-next this week with no reported issues\"\n\n* tag \u0027usb-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:\n usb: typec: tcpm: Skip sink_cap query only when VDM sm is busy\n usb: dwc3: gadget: Prevent EP queuing while stopping transfers\n usb: typec: tcpm: Invoke power_supply_changed for tcpm-source-psy-\n usb: typec: Remove vdo[3] part of tps6598x_rx_identity_reg struct\n usb-storage: Add quirk to defeat Kindle\u0027s automatic unload\n usb: gadget: configfs: Fix KASAN use-after-free\n usbip: Fix incorrect double assignment to udc-\u003eud.tcp_rx\n usb: cdnsp: Fixes incorrect value in ISOC TRB\n thunderbolt: Increase runtime PM reference count on DP tunnel discovery\n thunderbolt: Initialize HopID IDAs in tb_switch_alloc()\n" }, { "commit": "5ee96fa9dd78b2dc81b587e33074d877002d7605", "tree": "e87a3f76c5453f893b81db1eaee621b36ea84bfb", "parents": [ "1c74516c2da40ead2c22c4ac07c117c20f9c31ec", "81e2073c175b887398e5bca6c004efa89983f58d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:34:24 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:34:24 2021 -0700" }, "message": "Merge tag \u0027irq-urgent-2021-03-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull irq fix from Ingo Molnar:\n \"A change to robustify force-threaded IRQ handlers to always disable\n interrupts, plus a DocBook fix.\n\n The force-threaded IRQ handler change has been accelerated from the\n normal schedule of such a change to keep the bad pattern/workaround of\n spin_lock_irqsave() in handlers or IRQF_NOTHREAD as a kludge from\n spreading\"\n\n* tag \u0027irq-urgent-2021-03-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n genirq: Disable interrupts for force threaded handlers\n genirq/irq_sim: Fix typos in kernel doc (fnode -\u003e fwnode)\n" }, { "commit": "1c74516c2da40ead2c22c4ac07c117c20f9c31ec", "tree": "bb5ae2c31540088f66fd1be64b864a38d8ae8faf", "parents": [ "5ba33b488a04a13268131b0b4748a7c6f3598693", "2dc0572f2cef87425147658698dce2600b799bd3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:26:21 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:26:21 2021 -0700" }, "message": "Merge tag \u0027perf-urgent-2021-03-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull perf fixes from Ingo Molnar:\n \"Boundary condition fixes for bugs unearthed by the perf fuzzer\"\n\n* tag \u0027perf-urgent-2021-03-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n perf/x86/intel: Fix unchecked MSR access error caused by VLBR_EVENT\n perf/x86/intel: Fix a crash caused by zero PEBS status\n" }, { "commit": "5ba33b488a04a13268131b0b4748a7c6f3598693", "tree": "a73c8b667a027d6c0febf2ca3c4cfb1c6e7524d5", "parents": [ "92ed88cb4d7cd01eadb0f71063ceb5202548ce35", "38c93587375053c5b9ef093f4a5ea754538cba32" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:19:29 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:19:29 2021 -0700" }, "message": "Merge tag \u0027locking-urgent-2021-03-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull locking fixes from Ingo Molnar:\n\n - Get static calls \u0026 modules right. Hopefully.\n\n - WW mutex fixes\n\n* tag \u0027locking-urgent-2021-03-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n static_call: Fix static_call_update() sanity check\n static_call: Align static_call_is_init() patching condition\n static_call: Fix static_call_set_init()\n locking/ww_mutex: Fix acquire/release imbalance in ww_acquire_init()/ww_acquire_fini()\n locking/ww_mutex: Simplify use_ww_ctx \u0026 ww_ctx handling\n" }, { "commit": "92ed88cb4d7cd01eadb0f71063ceb5202548ce35", "tree": "11a3aad72e92ef3f68d8c3a9d33f7e86165276c9", "parents": [ "5e3ddf96e75983e4c404467fbb61b92d09333a1f", "429257a430a0e81e9979256e0db718e35e7d9cee" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:11:22 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:11:22 2021 -0700" }, "message": "Merge tag \u0027efi-urgent-2021-03-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull EFI fixes from Ingo Molnar:\n\n - another missing RT_PROP table related fix, to ensure that the\n efivarfs pseudo filesystem fails gracefully if variable services\n are unsupported\n\n - use the correct alignment for literal EFI GUIDs\n\n - fix a use after unmap issue in the memreserve code\n\n* tag \u0027efi-urgent-2021-03-21\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n efi: use 32-bit alignment for efi_guid_t literals\n firmware/efi: Fix a use after bug in efi_mem_reserve_persistent\n efivars: respect EFI_UNSUPPORTED return from firmware\n" }, { "commit": "5e3ddf96e75983e4c404467fbb61b92d09333a1f", "tree": "f45a21d97aa4db1372c599e279c099c775fae3bc", "parents": [ "b35660a7cebdf438e01bba05075ae2bcc0125650", "dd926880da8dbbe409e709c1d3c1620729a94732" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:04:20 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 11:04:20 2021 -0700" }, "message": "Merge tag \u0027x86_urgent_for_v5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86 fixes from Borislav Petkov:\n \"The freshest pile of shiny x86 fixes for 5.12:\n\n - Add the arch-specific mapping between physical and logical CPUs to\n fix devicetree-node lookups\n\n - Restore the IRQ2 ignore logic\n\n - Fix get_nr_restart_syscall() to return the correct restart syscall\n number. Split in a 4-patches set to avoid kABI breakage when\n backporting to dead kernels\"\n\n* tag \u0027x86_urgent_for_v5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n x86/apic/of: Fix CPU devicetree-node lookups\n x86/ioapic: Ignore IRQ2 again\n x86: Introduce restart_block-\u003earch_data to remove TS_COMPAT_RESTART\n x86: Introduce TS_COMPAT_RESTART to fix get_nr_restart_syscall()\n x86: Move TS_COMPAT back to asm/thread_info.h\n kernel, fs: Introduce and use set_restart_fn() and arch_set_restart_data()\n" }, { "commit": "b35660a7cebdf438e01bba05075ae2bcc0125650", "tree": "4f481813702aaea980735159d2aef4bb699adab6", "parents": [ "812da4d39463a060738008a46cfc9f775e4bfcf6", "cc7a0bb058b85ea03db87169c60c7cfdd5d34678" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 10:57:35 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 21 10:57:35 2021 -0700" }, "message": "Merge tag \u0027powerpc-5.12-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux\n\nPull powerpc fixes from Michael Ellerman:\n\n - Fix a possible stack corruption and subsequent DLPAR failure in the\n rpadlpar_io PCI hotplug driver\n\n - Two build fixes for uncommon configurations\n\nThanks to Christophe Leroy and Tyrel Datwyler.\n\n* tag \u0027powerpc-5.12-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:\n PCI: rpadlpar: Fix potential drc_name corruption in store functions\n powerpc: Force inlining of cpu_has_feature() to avoid build failure\n powerpc/vdso32: Add missing _restgpr_31_x to fix build failure\n" }, { "commit": "0031275d119efe16711cd93519b595e6f9b4b330", "tree": "c10e02cdb624d018a3120f91e1eaed2692bb0689", "parents": [ "00ddff431a458bbf143ea7c4c42d022676da1b17" ], "author": { "name": "Stefan Metzmacher", "email": "metze@samba.org", "time": "Sat Mar 20 20:33:36 2021 +0100" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Sun Mar 21 09:41:14 2021 -0600" }, "message": "io_uring: call req_set_fail_links() on short send[msg]()/recv[msg]() with MSG_WAITALL\n\nWithout that it\u0027s not safe to use them in a linked combination with\nothers.\n\nNow combinations like IORING_OP_SENDMSG followed by IORING_OP_SPLICE\nshould be possible.\n\nWe already handle short reads and writes for the following opcodes:\n\n- IORING_OP_READV\n- IORING_OP_READ_FIXED\n- IORING_OP_READ\n- IORING_OP_WRITEV\n- IORING_OP_WRITE_FIXED\n- IORING_OP_WRITE\n- IORING_OP_SPLICE\n- IORING_OP_TEE\n\nNow we have it for these as well:\n\n- IORING_OP_SENDMSG\n- IORING_OP_SEND\n- IORING_OP_RECVMSG\n- IORING_OP_RECV\n\nFor IORING_OP_RECVMSG we also check for the MSG_TRUNC and MSG_CTRUNC\nflags in order to call req_set_fail_links().\n\nThere might be applications arround depending on the behavior\nthat even short send[msg]()/recv[msg]() retuns continue an\nIOSQE_IO_LINK chain.\n\nIt\u0027s very unlikely that such applications pass in MSG_WAITALL,\nwhich is only defined in \u0027man 2 recvmsg\u0027, but not in \u0027man 2 sendmsg\u0027.\n\nIt\u0027s expected that the low level sock_sendmsg() call just ignores\nMSG_WAITALL, as MSG_ZEROCOPY is also ignored without explicitly set\nSO_ZEROCOPY.\n\nWe also expect the caller to know about the implicit truncation to\nMAX_RW_COUNT, which we don\u0027t detect.\n\ncc: netdev@vger.kernel.org\nLink: https://lore.kernel.org/r/c4e1a4cc0d905314f4d5dc567e65a7b09621aab3.1615908477.git.metze@samba.org\nSigned-off-by: Stefan Metzmacher \u003cmetze@samba.org\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "00ddff431a458bbf143ea7c4c42d022676da1b17", "tree": "a2c094c16df92bbe8c2040d7346541d019b1b2be", "parents": [ "4db4b1a0d1779dc159f7b87feb97030ec0b12597" ], "author": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Sun Mar 21 07:06:56 2021 -0600" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Sun Mar 21 09:41:14 2021 -0600" }, "message": "io-wq: ensure task is running before processing task_work\n\nMark the current task as running if we need to run task_work from the\nio-wq threads as part of work handling. If that is the case, then return\nas such so that the caller can appropriately loop back and reset if it\nwas part of a going-to-sleep flush.\n\nFixes: 3bfe6106693b (\"io-wq: fork worker threads from original task\")\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "4db4b1a0d1779dc159f7b87feb97030ec0b12597", "tree": "4a5fda2de8bfda249c99f791346d0f2faff9d4ed", "parents": [ "5be28c8f85ce99ed2d329d2ad8bdd18ea19473a5" ], "author": { "name": "Eric W. Biederman", "email": "ebiederm@xmission.com", "time": "Sun Mar 21 09:37:48 2021 -0600" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Sun Mar 21 09:41:07 2021 -0600" }, "message": "signal: don\u0027t allow STOP on PF_IO_WORKER threads\n\nJust like we don\u0027t allow normal signals to IO threads, don\u0027t deliver a\nSTOP to a task that has PF_IO_WORKER set. The IO threads don\u0027t take\nsignals in general, and have no means of flushing out a stop either.\n\nLonger term, we may want to look into allowing stop of these threads,\nas it relates to eg process freezing. For now, this prevents a spin\nissue if a SIGSTOP is delivered to the parent task.\n\nReported-by: Stefan Metzmacher \u003cmetze@samba.org\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\nSigned-off-by: \"Eric W. Biederman\" \u003cebiederm@xmission.com\u003e\n" }, { "commit": "5be28c8f85ce99ed2d329d2ad8bdd18ea19473a5", "tree": "a0e1f35dc49381c58cfb40e10d32d14aaf5996ed", "parents": [ "de75a3d3f5a14c9ab3c4883de3471d3c92a8ee78" ], "author": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Fri Mar 19 19:25:13 2021 -0600" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Sun Mar 21 09:39:32 2021 -0600" }, "message": "signal: don\u0027t allow sending any signals to PF_IO_WORKER threads\n\nThey don\u0027t take signals individually, and even if they share signals with\nthe parent task, don\u0027t allow them to be delivered through the worker\nthread. Linux does allow this kind of behavior for regular threads, but\nit\u0027s really a compatability thing that we need not care about for the IO\nthreads.\n\nReported-by: Stefan Metzmacher \u003cmetze@samba.org\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "64395d950bc476106b39341e42ebfd4d2eb71d2c", "tree": "ae3c8f192135909dd31d8d4b5c1a30361e779f90", "parents": [ "512c15ef05d73a04f1aef18a3bc61a8bb516f323" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Mar 21 00:45:37 2021 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Mar 21 00:45:37 2021 -0400" }, "message": "ext4: initialize ret to suppress smatch warning\n\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "512c15ef05d73a04f1aef18a3bc61a8bb516f323", "tree": "a07486905a2312bb0d0ba14019c01989a5d0780c", "parents": [ "8210bb29c1b66200cff7b25febcf6e39baf49fbf" ], "author": { "name": "Pan Bian", "email": "bianpan2016@163.com", "time": "Sun Jan 17 00:57:32 2021 -0800" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Mar 21 00:42:12 2021 -0400" }, "message": "ext4: stop inode update before return\n\nThe inode update should be stopped before returing the error code.\n\nSigned-off-by: Pan Bian \u003cbianpan2016@163.com\u003e\nLink: https://lore.kernel.org/r/20210117085732.93788-1-bianpan2016@163.com\nFixes: 8016e29f4362 (\"ext4: fast commit recovery path\")\nCc: stable@kernel.org\nReviewed-by: Harshad Shirwadkar \u003charshadshirwadkar@gmail.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "8210bb29c1b66200cff7b25febcf6e39baf49fbf", "tree": "9f3f60f830d67c9fede42c5b29b051e9f075269a", "parents": [ "2a4ae3bcdf05b8639406eaa09a2939f3c6dd8e75" ], "author": { "name": "Harshad Shirwadkar", "email": "harshadshirwadkar@gmail.com", "time": "Tue Mar 16 15:19:21 2021 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Mar 21 00:38:18 2021 -0400" }, "message": "ext4: fix rename whiteout with fast commit\n\nThis patch adds rename whiteout support in fast commits. Note that the\nwhiteout object that gets created is actually char device. Which\nimples, the function ext4_inode_journal_mode(struct inode *inode)\nwould return \"JOURNAL_DATA\" for this inode. This has a consequence in\nfast commit code that it will make creation of the whiteout object a\nfast-commit ineligible behavior and thus will fall back to full\ncommits. With this patch, this can be observed by running fast commits\nwith rename whiteout and seeing the stats generated by ext4_fc_stats\ntracepoint as follows:\n\next4_fc_stats: dev 254:32 fc ineligible reasons:\nXATTR:0, CROSS_RENAME:0, JOURNAL_FLAG_CHANGE:0, NO_MEM:0, SWAP_BOOT:0,\nRESIZE:0, RENAME_DIR:0, FALLOC_RANGE:0, INODE_JOURNAL_DATA:16;\nnum_commits:6, ineligible: 6, numblks: 3\n\nSo in short, this patch guarantees that in case of rename whiteout, we\nfall back to full commits.\n\nAmir mentioned that instead of creating a new whiteout object for\nevery rename, we can create a static whiteout object with irrelevant\nnlink. That will make fast commits to not fall back to full\ncommit. But until this happens, this patch will ensure correctness by\nfalling back to full commits.\n\nFixes: 8016e29f4362 (\"ext4: fast commit recovery path\")\nCc: stable@kernel.org\nSigned-off-by: Harshad Shirwadkar \u003charshadshirwadkar@gmail.com\u003e\nLink: https://lore.kernel.org/r/20210316221921.1124955-1-harshadshirwadkar@gmail.com\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "2a4ae3bcdf05b8639406eaa09a2939f3c6dd8e75", "tree": "91b8caa387423ca6b69fa991141da04dd739eb58", "parents": [ "7d8bd3c76da1d94b85e6c9b7007e20e980bfcfe6" ], "author": { "name": "Jan Kara", "email": "jack@suse.cz", "time": "Mon Mar 15 17:59:06 2021 +0100" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Mar 21 00:27:49 2021 -0400" }, "message": "ext4: fix timer use-after-free on failed mount\n\nWhen filesystem mount fails because of corrupted filesystem we first\ncancel the s_err_report timer reminding fs errors every day and only\nthen we flush s_error_work. However s_error_work may report another fs\nerror and re-arm timer thus resulting in timer use-after-free. Fix the\nproblem by first flushing the work and only after that canceling the\ns_err_report timer.\n\nReported-by: syzbot+628472a2aac693ab0fcd@syzkaller.appspotmail.com\nFixes: 2d01ddc86606 (\"ext4: save error info to sb through journal if available\")\nCC: stable@vger.kernel.org\nSigned-off-by: Jan Kara \u003cjack@suse.cz\u003e\nLink: https://lore.kernel.org/r/20210315165906.2175-1-jack@suse.cz\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "7d8bd3c76da1d94b85e6c9b7007e20e980bfcfe6", "tree": "0cee9b98391bae8e7c9f5c94a13453258f58b122", "parents": [ "6b22489911b726eebbf169caee52fea52013fbdd" ], "author": { "name": "Shijie Luo", "email": "luoshijie1@huawei.com", "time": "Fri Mar 12 01:50:51 2021 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Mar 21 00:14:08 2021 -0400" }, "message": "ext4: fix potential error in ext4_do_update_inode\n\nIf set_large_file \u003d 1 and errors occur in ext4_handle_dirty_metadata(),\nthe error code will be overridden, go to out_brelse to avoid this\nsituation.\n\nSigned-off-by: Shijie Luo \u003cluoshijie1@huawei.com\u003e\nLink: https://lore.kernel.org/r/20210312065051.36314-1-luoshijie1@huawei.com\nCc: stable@kernel.org\nReviewed-by: Jan Kara \u003cjack@suse.cz\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "6b22489911b726eebbf169caee52fea52013fbdd", "tree": "244fdd3d130ae305d1f672faa18b919f3905a71f", "parents": [ "5dccdc5a1916d4266edd251f20bbbb113a5c495f" ], "author": { "name": "zhangyi (F)", "email": "yi.zhang@huawei.com", "time": "Fri Mar 05 20:05:08 2021 +0800" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Mar 21 00:09:17 2021 -0400" }, "message": "ext4: do not try to set xattr into ea_inode if value is empty\n\nSyzbot report a warning that ext4 may create an empty ea_inode if set\nan empty extent attribute to a file on the file system which is no free\nblocks left.\n\n WARNING: CPU: 6 PID: 10667 at fs/ext4/xattr.c:1640 ext4_xattr_set_entry+0x10f8/0x1114 fs/ext4/xattr.c:1640\n ...\n Call trace:\n ext4_xattr_set_entry+0x10f8/0x1114 fs/ext4/xattr.c:1640\n ext4_xattr_block_set+0x1d0/0x1b1c fs/ext4/xattr.c:1942\n ext4_xattr_set_handle+0x8a0/0xf1c fs/ext4/xattr.c:2390\n ext4_xattr_set+0x120/0x1f0 fs/ext4/xattr.c:2491\n ext4_xattr_trusted_set+0x48/0x5c fs/ext4/xattr_trusted.c:37\n __vfs_setxattr+0x208/0x23c fs/xattr.c:177\n ...\n\nNow, ext4 try to store extent attribute into an external inode if\next4_xattr_block_set() return -ENOSPC, but for the case of store an\nempty extent attribute, store the extent entry into the extent\nattribute block is enough. A simple reproduce below.\n\n fallocate test.img -l 1M\n mkfs.ext4 -F -b 2048 -O ea_inode test.img\n mount test.img /mnt\n dd if\u003d/dev/zero of\u003d/mnt/foo bs\u003d2048 count\u003d500\n setfattr -n \"user.test\" /mnt/foo\n\nReported-by: syzbot+98b881fdd8ebf45ab4ae@syzkaller.appspotmail.com\nFixes: 9c6e7853c531 (\"ext4: reserve space for xattr entries/names\")\nCc: stable@kernel.org\nSigned-off-by: zhangyi (F) \u003cyi.zhang@huawei.com\u003e\nLink: https://lore.kernel.org/r/20210305120508.298465-1-yi.zhang@huawei.com\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "5dccdc5a1916d4266edd251f20bbbb113a5c495f", "tree": "331426781ad251b0c88a4c35e69430a536eb60ff", "parents": [ "b7ff91fd030dc9d72ed91b1aab36e445a003af4f" ], "author": { "name": "zhangyi (F)", "email": "yi.zhang@huawei.com", "time": "Wed Mar 03 21:17:03 2021 +0800" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Mar 21 00:09:14 2021 -0400" }, "message": "ext4: do not iput inode under running transaction in ext4_rename()\n\nIn ext4_rename(), when RENAME_WHITEOUT failed to add new entry into\ndirectory, it ends up dropping new created whiteout inode under the\nrunning transaction. After commit \u003c9b88f9fb0d2\u003e (\"ext4: Do not iput inode\nunder running transaction\"), we follow the assumptions that evict() does\nnot get called from a transaction context but in ext4_rename() it breaks\nthis suggestion. Although it\u0027s not a real problem, better to obey it, so\nthis patch add inode to orphan list and stop transaction before final\niput().\n\nSigned-off-by: zhangyi (F) \u003cyi.zhang@huawei.com\u003e\nLink: https://lore.kernel.org/r/20210303131703.330415-2-yi.zhang@huawei.com\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "b7ff91fd030dc9d72ed91b1aab36e445a003af4f", "tree": "7015e62beaf3916d42cec195a5e137ec2e4c7ff6", "parents": [ "f053cf7aa66cd9d592b0fc967f4d887c2abff1b7" ], "author": { "name": "zhangyi (F)", "email": "yi.zhang@huawei.com", "time": "Wed Mar 03 21:17:02 2021 +0800" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Mar 21 00:03:39 2021 -0400" }, "message": "ext4: find old entry again if failed to rename whiteout\n\nIf we failed to add new entry on rename whiteout, we cannot reset the\nold-\u003ede entry directly, because the old-\u003ede could have moved from under\nus during make indexed dir. So find the old entry again before reset is\nneeded, otherwise it may corrupt the filesystem as below.\n\n /dev/sda: Entry \u002700000001\u0027 in ??? (12) has deleted/unused inode 15. CLEARED.\n /dev/sda: Unattached inode 75\n /dev/sda: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.\n\nFixes: 6b4b8e6b4ad (\"ext4: fix bug for rename with RENAME_WHITEOUT\")\nCc: stable@vger.kernel.org\nSigned-off-by: zhangyi (F) \u003cyi.zhang@huawei.com\u003e\nLink: https://lore.kernel.org/r/20210303131703.330415-1-yi.zhang@huawei.com\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "81e2073c175b887398e5bca6c004efa89983f58d", "tree": "67aa8a85a9e28345230493605c2465c21b870860", "parents": [ "ef4cb70a4c22bf301cd757dcc838dc8ca9526477" ], "author": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Wed Mar 17 15:38:52 2021 +0100" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Sun Mar 21 00:17:52 2021 +0100" }, "message": "genirq: Disable interrupts for force threaded handlers\n\nWith interrupt force threading all device interrupt handlers are invoked\nfrom kernel threads. Contrary to hard interrupt context the invocation only\ndisables bottom halfs, but not interrupts. This was an oversight back then\nbecause any code like this will have an issue:\n\nthread(irq_A)\n irq_handler(A)\n spin_lock(\u0026foo-\u003elock);\n\ninterrupt(irq_B)\n irq_handler(B)\n spin_lock(\u0026foo-\u003elock);\n\nThis has been triggered with networking (NAPI vs. hrtimers) and console\ndrivers where printk() happens from an interrupt which interrupted the\nforce threaded handler.\n\nNow people noticed and started to change the spin_lock() in the handler to\nspin_lock_irqsave() which affects performance or add IRQF_NOTHREAD to the\ninterrupt request which in turn breaks RT.\n\nFix the root cause and not the symptom and disable interrupts before\ninvoking the force threaded handler which preserves the regular semantics\nand the usefulness of the interrupt force threading as a general debugging\ntool.\n\nFor not RT this is not changing much, except that during the execution of\nthe threaded handler interrupts are delayed until the handler\nreturns. Vs. scheduling and softirq processing there is no difference.\n\nFor RT kernels there is no issue.\n\nFixes: 8d32a307e4fa (\"genirq: Provide forced interrupt threading\")\nReported-by: Johan Hovold \u003cjohan@kernel.org\u003e\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nReviewed-by: Johan Hovold \u003cjohan@kernel.org\u003e\nAcked-by: Sebastian Andrzej Siewior \u003cbigeasy@linutronix.de\u003e\nLink: https://lore.kernel.org/r/20210317143859.513307808@linutronix.de\n\n" }, { "commit": "812da4d39463a060738008a46cfc9f775e4bfcf6", "tree": "dc065a9ac51a10a4096a46394e257179f09bd169", "parents": [ "bfdc4aa9e99204d1bf925cdbfea9a1d903ebed29", "a5406a7ff56e63376c210b06072aa0ef23473366" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 20 11:01:54 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 20 11:01:54 2021 -0700" }, "message": "Merge tag \u0027riscv-for-linus-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux\n\nPull RISC-V fixes from Palmer Dabbelt:\n \"A handful of fixes for 5.12:\n\n - fix the SBI remote fence numbers for hypervisor fences, which had\n been transcribed in the wrong order in Linux. These fences are only\n used with the KVM patches applied.\n\n - fix a whole host of build warnings, these should have no functional\n change.\n\n - fix init_resources() to prevent an off-by-one error from causing an\n out-of-bounds array reference. This was manifesting during boot on\n vexriscv.\n\n - ensure the KASAN mappings are visible before proceeding to use\n them\"\n\n* tag \u0027riscv-for-linus-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:\n riscv: Correct SPARSEMEM configuration\n RISC-V: kasan: Declare kasan_shallow_populate() static\n riscv: Ensure page table writes are flushed when initializing KASAN vmalloc\n RISC-V: Fix out-of-bounds accesses in init_resources()\n riscv: Fix compilation error with Canaan SoC\n ftrace: Fix spelling mistake \"disabed\" -\u003e \"disabled\"\n riscv: fix bugon.cocci warnings\n riscv: process: Fix no prototype for arch_dup_task_struct\n riscv: ftrace: Use ftrace_get_regs helper\n riscv: process: Fix no prototype for show_regs\n riscv: syscall_table: Reduce W\u003d1 compilation warnings noise\n riscv: time: Fix no prototype for time_init\n riscv: ptrace: Fix no prototype warnings\n riscv: sbi: Fix comment of __sbi_set_timer_v01\n riscv: irq: Fix no prototype warning\n riscv: traps: Fix no prototype warnings\n RISC-V: correct enum sbi_ext_rfence_fid\n" }, { "commit": "bfdc4aa9e99204d1bf925cdbfea9a1d903ebed29", "tree": "a42605c351d575d51e7d1b6da335ed67d7052bc0", "parents": [ "af97713dff9f877922af35f0796e1d76b8a4be00", "65af8f0166f4d15e61c63db498ec7981acdd897f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 20 11:00:25 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 20 11:00:25 2021 -0700" }, "message": "Merge tag \u00275.12-rc3-smb3\u0027 of git://git.samba.org/sfrench/cifs-2.6\n\nPull cifs fixes from Steve French:\n \"Five cifs/smb3 fixes - three for stable, including an important ACL\n fix and security signature fix\"\n\n* tag \u00275.12-rc3-smb3\u0027 of git://git.samba.org/sfrench/cifs-2.6:\n cifs: fix allocation size on newly created files\n cifs: warn and fail if trying to use rootfs without the config option\n fs/cifs/: fix misspellings using codespell tool\n cifs: Fix preauth hash corruption\n cifs: update new ACE pointer after populate_new_aces.\n" }, { "commit": "af97713dff9f877922af35f0796e1d76b8a4be00", "tree": "0ef949e6bc1780d73a2bf86eda7dc781c26dd7d1", "parents": [ "1c273e10bc0cc7efb933e0ca10e260cdfc9f0b8c", "a50bd64616907ed126ffbdbaa06c5ce708c4a404" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 20 10:57:10 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 20 10:57:10 2021 -0700" }, "message": "Merge tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi\n\nPull SCSI fixes from James Bottomley:\n \"Eight fixes, all in drivers, all fairly minor either being fixes in\n error legs, memory leaks on teardown, context errors or semantic\n problems\"\n\n* tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:\n scsi: mpt3sas: Do not use GFP_KERNEL in atomic context\n scsi: ufs: ufs-mediatek: Correct operator \u0026 -\u003e \u0026\u0026\n scsi: sd_zbc: Update write pointer offset cache\n scsi: lpfc: Fix some error codes in debugfs\n scsi: qla2xxx: Fix broken #endif placement\n scsi: st: Fix a use after free in st_open()\n scsi: myrs: Fix a double free in myrs_cleanup()\n scsi: ibmvfc: Free channel_setup_buf during device tear down\n" }, { "commit": "1c273e10bc0cc7efb933e0ca10e260cdfc9f0b8c", "tree": "9d1e3522ba86430c9f056ee605c4748cf0211441", "parents": [ "d626c692aaeb2ff839bfe463f096660c39a6d1eb", "6980d29ce4da223ad7f0751c7f1d61d3c6b54ab3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 17:32:30 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 17:32:30 2021 -0700" }, "message": "Merge tag \u0027zonefs-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/zonefs\n\nPull zonefs fixes from Damien Le Moal:\n\n - fix inode write open reference count (Chao)\n\n - Fix wrong write offset for asynchronous O_APPEND writes (me)\n\n - Prevent use of sequential zone file as swap files (me)\n\n* tag \u0027zonefs-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/zonefs:\n zonefs: fix to update .i_wr_refcnt correctly in zonefs_open_zone()\n zonefs: Fix O_APPEND async write handling\n zonefs: prevent use of seq files as swap file\n" }, { "commit": "d626c692aaeb2ff839bfe463f096660c39a6d1eb", "tree": "863a5b41c2f2205d4d689565bf2132e08c944fc3", "parents": [ "0ada2dad8bf39857f25e6ecbf68bb1664ca1ee5b", "d38b4d289486daee01c1fdf056b46b7cdfe72e9e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 17:07:10 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 17:07:10 2021 -0700" }, "message": "Merge tag \u0027block-5.12-2021-03-19\u0027 of git://git.kernel.dk/linux-block\n\nPull block fixes from Jens Axboe:\n \"Just an NVMe pull request this week:\n\n - fix tag allocation for keep alive\n\n - fix a unit mismatch for the Write Zeroes limits\n\n - various TCP transport fixes (Sagi Grimberg, Elad Grupi)\n\n - fix iosqes and iocqes validation for discovery controllers (Sagi Grimberg)\"\n\n* tag \u0027block-5.12-2021-03-19\u0027 of git://git.kernel.dk/linux-block:\n nvmet-tcp: fix kmap leak when data digest in use\n nvmet: don\u0027t check iosqes,iocqes for discovery controllers\n nvme-rdma: fix possible hang when failing to set io queues\n nvme-tcp: fix possible hang when failing to set io queues\n nvme-tcp: fix misuse of __smp_processor_id with preemption enabled\n nvme-tcp: fix a NULL deref when receiving a 0-length r2t PDU\n nvme: fix Write Zeroes limitations\n nvme: allocate the keep alive request using BLK_MQ_REQ_NOWAIT\n nvme: merge nvme_keep_alive into nvme_keep_alive_work\n nvme-fabrics: only reserve a single tag\n" }, { "commit": "0ada2dad8bf39857f25e6ecbf68bb1664ca1ee5b", "tree": "b5023b73d0f482f594cfa82f374c7d0081a1249a", "parents": [ "ecd8ee7f9c1af253738ca4321509ddee727d468d", "de75a3d3f5a14c9ab3c4883de3471d3c92a8ee78" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 17:01:09 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 17:01:09 2021 -0700" }, "message": "Merge tag \u0027io_uring-5.12-2021-03-19\u0027 of git://git.kernel.dk/linux-block\n\nPull io_uring fixes from Jens Axboe:\n \"Quieter week this time, which was both expected and desired. About\n half of the below is fixes for this release, the other half are just\n fixes in general. In detail:\n\n - Fix the freezing of IO threads, by making the freezer not send them\n fake signals. Make them freezable by default.\n\n - Like we did for personalities, move the buffer IDR to xarray. Kills\n some code and avoids a use-after-free on teardown.\n\n - SQPOLL cleanups and fixes (Pavel)\n\n - Fix linked timeout race (Pavel)\n\n - Fix potential completion post use-after-free (Pavel)\n\n - Cleanup and move internal structures outside of general kernel view\n (Stefan)\n\n - Use MSG_SIGNAL for send/recv from io_uring (Stefan)\"\n\n* tag \u0027io_uring-5.12-2021-03-19\u0027 of git://git.kernel.dk/linux-block:\n io_uring: don\u0027t leak creds on SQO attach error\n io_uring: use typesafe pointers in io_uring_task\n io_uring: remove structures from include/linux/io_uring.h\n io_uring: imply MSG_NOSIGNAL for send[msg]()/recv[msg]() calls\n io_uring: fix sqpoll cancellation via task_work\n io_uring: add generic callback_head helpers\n io_uring: fix concurrent parking\n io_uring: halt SQO submission on ctx exit\n io_uring: replace sqd rw_semaphore with mutex\n io_uring: fix complete_post use ctx after free\n io_uring: fix -\u003eflags races by linked timeouts\n io_uring: convert io_buffer_idr to XArray\n io_uring: allow IO worker threads to be frozen\n kernel: freezer should treat PF_IO_WORKER like PF_KTHREAD for freezing\n" }, { "commit": "dd926880da8dbbe409e709c1d3c1620729a94732", "tree": "7ba8c1bd2c8853046bc97a51ec39b7c0f85bfa3a", "parents": [ "a501b048a95b79e1e34f03cac3c87ff1e9f229ad" ], "author": { "name": "Johan Hovold", "email": "johan@kernel.org", "time": "Fri Mar 12 10:20:33 2021 +0100" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Fri Mar 19 23:01:49 2021 +0100" }, "message": "x86/apic/of: Fix CPU devicetree-node lookups\n\nArchitectures that describe the CPU topology in devicetree and do not have\nan identity mapping between physical and logical CPU ids must override the\ndefault implementation of arch_match_cpu_phys_id().\n\nFailing to do so breaks CPU devicetree-node lookups using of_get_cpu_node()\nand of_cpu_device_node_get() which several drivers rely on. It also causes\nthe CPU struct devices exported through sysfs to point to the wrong\ndevicetree nodes.\n\nOn x86, CPUs are described in devicetree using their APIC ids and those\ndo not generally coincide with the logical ids, even if CPU0 typically\nuses APIC id 0.\n\nAdd the missing implementation of arch_match_cpu_phys_id() so that CPU-node\nlookups work also with SMP.\n\nApart from fixing the broken sysfs devicetree-node links this likely does\nnot affect current users of mainline kernels on x86.\n\nFixes: 4e07db9c8db8 (\"x86/devicetree: Use CPU description from Device Tree\")\nSigned-off-by: Johan Hovold \u003cjohan@kernel.org\u003e\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nLink: https://lore.kernel.org/r/20210312092033.26317-1-johan@kernel.org\n" }, { "commit": "ecd8ee7f9c1af253738ca4321509ddee727d468d", "tree": "b683a542f68c52000db681b4886c87c14fea7580", "parents": [ "3149860dc717e8dd339d89d17ebe615cb09e158b", "9ce3746d64132a561bceab6421715e7c04e85074" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 14:10:07 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 14:10:07 2021 -0700" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm\n\nPull kvm fixes from Paolo Bonzini:\n \"Fixes for kvm on x86:\n\n - new selftests\n\n - fixes for migration with HyperV re-enlightenment enabled\n\n - fix RCU/SRCU usage\n\n - fixes for local_irq_restore misuse false positive\"\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm:\n documentation/kvm: additional explanations on KVM_SET_BOOT_CPU_ID\n x86/kvm: Fix broken irq restoration in kvm_wait\n KVM: X86: Fix missing local pCPU when executing wbinvd on all dirty pCPUs\n KVM: x86: Protect userspace MSR filter with SRCU, and set atomically-ish\n selftests: kvm: add set_boot_cpu_id test\n selftests: kvm: add _vm_ioctl\n selftests: kvm: add get_msr_index_features\n selftests: kvm: Add basic Hyper-V clocksources tests\n KVM: x86: hyper-v: Don\u0027t touch TSC page values when guest opted for re-enlightenment\n KVM: x86: hyper-v: Track Hyper-V TSC page status\n KVM: x86: hyper-v: Prevent using not-yet-updated TSC page by secondary CPUs\n KVM: x86: hyper-v: Limit guest to writing zero to HV_X64_MSR_TSC_EMULATION_STATUS\n KVM: x86/mmu: Store the address space ID in the TDP iterator\n KVM: x86/mmu: Factor out tdp_iter_return_to_root\n KVM: x86/mmu: Fix RCU usage when atomically zapping SPTEs\n KVM: x86/mmu: Fix RCU usage in handle_removed_tdp_mmu_page\n" }, { "commit": "3149860dc717e8dd339d89d17ebe615cb09e158b", "tree": "7ca48b7d09ec351dc43bf9adf3f19d64da0d7634", "parents": [ "6bfea141b3d26898705704efd18401d91afcbb0a", "6cb59afe9e5b45a035bd6b97da6593743feefc72" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 14:07:19 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 14:07:19 2021 -0700" }, "message": "Merge tag \u0027gpio-fixes-for-v5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux\n\nPull gpio fixes from Bartosz Golaszewski:\n \"Two fixes for the GPIO subsystem. Both address issues in the core GPIO\n code:\n\n - fix the return value in error path in gpiolib_dev_init()\n\n - fix the \u0027gpio-line-names\u0027 property handling correctly this time\"\n\n* tag \u0027gpio-fixes-for-v5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:\n gpiolib: Assign fwnode to parent\u0027s if no primary one provided\n gpiolib: Fix error return code in gpiolib_dev_init()\n" }, { "commit": "6bfea141b3d26898705704efd18401d91afcbb0a", "tree": "12f49961ab4fab5d4faeab0938e914f89cf0dff4", "parents": [ "278924cb99c93861c1cc3d266d719095bbd84f16", "0b13525c20febcfecccf6fc1db5969727401317d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 11:39:28 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 11:39:28 2021 -0700" }, "message": "Merge tag \u0027s390-5.12-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux\n\nPull s390 updates from Heiko Carstens:\n\n - disable preemption when accessing local per-cpu variables in the new\n counter set driver\n\n - fix by a factor of four increased steal time due to missing\n cputime_to_nsecs() conversion\n\n - fix PCI device structure leak\n\n* tag \u0027s390-5.12-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:\n s390/pci: fix leak of PCI device structure\n s390/vtime: fix increased steal time accounting\n s390/cpumf: disable preemption when accessing per-cpu variable\n" }, { "commit": "278924cb99c93861c1cc3d266d719095bbd84f16", "tree": "fdc45f24e8b4ff56681563546982f76a4a007b51", "parents": [ "ec85720933863015b1c26bc19cf4e044da139bc5", "83b62687a05205847d627f29126a8fee3c644335" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 10:06:30 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 10:06:30 2021 -0700" }, "message": "Merge tag \u0027trace-v5.12-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace\n\nPull workqueue tracing fix from Steven Rostedt:\n \"Fix workqueue trace event unsafe string reference\n\n After adding a verifier to test all strings printed in trace events to\n make sure they either point to a string on the ring buffer, or to read\n only core kernel memory, it triggered on a workqueue trace event. The\n trace event workqueue_queue_work references the allocated name of the\n workqueue in the output. If the workqueue is freed before the trace is\n read, then the trace will dereference freed memory.\n\n Update the trace event to use the __string(), __assign_str(), and\n __get_str() helpers to handle such cases\"\n\n* tag \u0027trace-v5.12-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:\n workqueue/tracing: Copy workqueue name to buffer in trace event\n" }, { "commit": "ec85720933863015b1c26bc19cf4e044da139bc5", "tree": "cdc44a17f0d08bb00538300a45da67b717499b00", "parents": [ "65a103747104368cb1ba6f097dcc3f85b1dcf86b", "49cb71a77ce760f20487c38f891aa3132bef782e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 10:00:10 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 10:00:10 2021 -0700" }, "message": "Merge tag \u0027pm-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull power management fixes from Rafael Wysocki:\n \"Revert two problematic commits.\n\n Specifics:\n\n - Revert ACPI PM commit that attempted to improve reboot handling on\n some systems, but it caused other systems to panic() during reboot\n (Josef Bacik)\n\n - Revert PM-runtime commit that attempted to improve the handling of\n suppliers during PM-runtime suspend of a consumer device, but it\n introduced a race condition potentially leading to unexpected\n behavior (Rafael Wysocki)\"\n\n* tag \u0027pm-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n Revert \"PM: runtime: Update device status before letting suppliers suspend\"\n Revert \"PM: ACPI: reboot: Use S5 for reboot\"\n" }, { "commit": "65a103747104368cb1ba6f097dcc3f85b1dcf86b", "tree": "8f8d9ec52b987dc19b2699639a4b91cbb42a8ba8", "parents": [ "769e155c5395100fc468aa87703c486f276c16cd", "8dfd0fa6ecdc5e2099a57d485b7ce237abc6c7a0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 09:56:04 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 09:56:04 2021 -0700" }, "message": "Merge tag \u0027iommu-fixes-v5.12-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu\n\nPull iommu fixes from Joerg Roedel:\n\n - Three AMD IOMMU patches to fix a boot crash on AMD Stoney systems and\n every other AMD IOMMU system booted with \u0027amd_iommu\u003doff\u0027.\n\n This is a v5.11 regression.\n\n - A Fix for the Tegra IOMMU driver to make sure it detects all IOMMUs\n\n* tag \u0027iommu-fixes-v5.12-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:\n iommu/tegra-smmu: Make tegra_smmu_probe_device() to handle all IOMMU phandles\n iommu/amd: Keep track of amd_iommu_irq_remap state\n iommu/amd: Don\u0027t call early_amd_iommu_init() when AMD IOMMU is disabled\n iommu/amd: Move Stoney Ridge check to detect_ivrs()\n" }, { "commit": "769e155c5395100fc468aa87703c486f276c16cd", "tree": "8af16e9c671480e6d6a4eba183798c11b9f45238", "parents": [ "8b12a62a4e3ed4ae99c715034f557eb391d6b196", "50b1affc891cbc103a2334ce909a026e25f4c84d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 09:53:32 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 19 09:53:32 2021 -0700" }, "message": "Merge tag \u0027sound-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound\n\nPull sound fixes from Takashi Iwai:\n \"The majority of changes are various ASoC device/platform-specific\n small fixes (including a removal of stale file) while the only common\n change is a clk management fix in ASoC simple-card driver.\n\n The rest are the usual HD-audio quirks\"\n\n* tag \u0027sound-5.12-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound: (44 commits)\n ALSA: usb-audio: Fix unintentional sign extension issue\n ALSA: hda/realtek: fix mute/micmute LEDs for HP 850 G8\n ASoC: dt-bindings: fsl_spdif: Add compatible string for new platforms\n ASoC: rt711: add snd_soc_component remove callback\n ASoC: rt5659: Update MCLK rate in set_sysclk()\n ASoC: simple-card-utils: Do not handle device clock\n ALSA: hda/realtek: fix mute/micmute LEDs for HP 440 G8\n ALSA: hda/realtek: fix mute/micmute LEDs for HP 840 G8\n ALSA: hda/realtek: apply pin quirk for XiaomiNotebook Pro\n ALSA: hda/realtek: Apply headset-mic quirks for Xiaomi Redmibook Air\n ASoC: mediatek: mt8192: fix tdm out data is valid on rising edge\n ALSA: dice: fix null pointer dereference when node is disconnected\n ALSA: hda: generic: Fix the micmute led init state\n ASoC: qcom: lpass-cpu: Fix lpass dai ids parse\n spi: cadence: set cqspi to the driver_data field of struct device\n ASoC: SOF: intel: fix wrong poll bits in dsp power down\n ASoC: codecs: wcd934x: add a sanity check in set channel map\n ASoC: qcom: sdm845: Fix array out of range on rx slim channels\n ASoC: qcom: sdm845: Fix array out of bounds access\n ASoC: remove remnants of sirf prima/atlas audio codec\n ...\n" }, { "commit": "65af8f0166f4d15e61c63db498ec7981acdd897f", "tree": "529634cc41ad29c4136a7f31a43722341919eb90", "parents": [ "af3ef3b1031634724a3763606695ebcd113d782b" ], "author": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Fri Mar 19 00:05:48 2021 -0500" }, "committer": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Fri Mar 19 11:51:31 2021 -0500" }, "message": "cifs: fix allocation size on newly created files\n\nApplications that create and extend and write to a file do not\nexpect to see 0 allocation size. When file is extended,\nset its allocation size to a plausible value until we have a\nchance to query the server for it. When the file is cached\nthis will prevent showing an impossible number of allocated\nblocks (like 0). This fixes e.g. xfstests 614 which does\n\n 1) create a file and set its size to 64K\n 2) mmap write 64K to the file\n 3) stat -c %b for the file (to query the number of allocated blocks)\n\nIt was failing because we returned 0 blocks. Even though we would\nreturn the correct cached file size, we returned an impossible\nallocation size.\n\nSigned-off-by: Steve French \u003cstfrench@microsoft.com\u003e\nCC: \u003cstable@vger.kernel.org\u003e\nReviewed-by: Aurelien Aptel \u003caaptel@suse.com\u003e\n" }, { "commit": "49cb71a77ce760f20487c38f891aa3132bef782e", "tree": "55dd0221ec5cd04115faa8ecc8caae96ab0e1b3c", "parents": [ "9d3fcb28f9b9750b474811a2964ce022df56336e", "0cab893f409c53634d0d818fa414641cbcdb0dab" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Fri Mar 19 16:38:45 2021 +0100" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Fri Mar 19 16:38:45 2021 +0100" }, "message": "Merge branch \u0027pm-core\u0027\n\n* pm-core:\n Revert \"PM: runtime: Update device status before letting suppliers suspend\"\n" }, { "commit": "0cab893f409c53634d0d818fa414641cbcdb0dab", "tree": "e34ab8024ba7dc54187156289ca8cd9d12c8636e", "parents": [ "1e28eed17697bcf343c6743f0028cc3b5dd88bf0" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Fri Mar 19 15:47:25 2021 +0100" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Fri Mar 19 16:35:47 2021 +0100" }, "message": "Revert \"PM: runtime: Update device status before letting suppliers suspend\"\n\nRevert commit 44cc89f76464 (\"PM: runtime: Update device status\nbefore letting suppliers suspend\") that introduced a race condition\ninto __rpm_callback() which allowed a concurrent rpm_resume() to\nrun and resume the device prematurely after its status had been\nchanged to RPM_SUSPENDED by __rpm_callback().\n\nFixes: 44cc89f76464 (\"PM: runtime: Update device status before letting suppliers suspend\")\nLink: https://lore.kernel.org/linux-pm/24dfb6fc-5d54-6ee2-9195-26428b7ecf8a@intel.com/\nReported-by: Adrian Hunter \u003cadrian.hunter@intel.com\u003e\nCc: 4.10+ \u003cstable@vger.kernel.org\u003e # 4.10+\nSigned-off-by: Rafael J. Wysocki \u003crafael.j.wysocki@intel.com\u003e\nReviewed-by: Ulf Hansson \u003culf.hansson@linaro.org\u003e\n" }, { "commit": "429257a430a0e81e9979256e0db718e35e7d9cee", "tree": "bc8fa39608b41aa410457535a603bd78b598c79a", "parents": [ "1e28eed17697bcf343c6743f0028cc3b5dd88bf0", "fb98cc0b3af2ba4d87301dff2b381b12eee35d7d" ], "author": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Fri Mar 19 14:23:46 2021 +0100" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Fri Mar 19 14:23:46 2021 +0100" }, "message": "Merge tag \u0027efi-urgent-for-v5.12-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/efi/efi into efi/urgent\n\nPull EFI fixes from Ard Biesheuvel:\n\n \"- another missing RT_PROP table related fix, to ensure that the efivarfs\n pseudo filesystem fails gracefully if variable services are unsupported\n - use the correct alignment for literal EFI GUIDs\n - fix a use after unmap issue in the memreserve code\"\n\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "d38b4d289486daee01c1fdf056b46b7cdfe72e9e", "tree": "d40b2ebf6117bcfe97bad93a2d02d5ec46fda852", "parents": [ "1e28eed17697bcf343c6743f0028cc3b5dd88bf0", "bac04454ef9fada009f0572576837548b190bf94" ], "author": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Fri Mar 19 06:40:47 2021 -0600" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Fri Mar 19 06:40:47 2021 -0600" }, "message": "Merge tag \u0027nvme-5.12-20210319\u0027 of git://git.infradead.org/nvme into block-5.12\n\nPull NVMe updates from Christoph:\n\n\"nvme fixes for 5.12\n\n - fix tag allocation for keep alive\n - fix a unit mismatch for the Write Zeroes limits\n - various TCP transport fixes (Sagi Grimberg, Elad Grupi)\n - fix iosqes and iocqes validation for discovery controllers (Sagi Grimberg)\"\n\n* tag \u0027nvme-5.12-20210319\u0027 of git://git.infradead.org/nvme:\n nvmet-tcp: fix kmap leak when data digest in use\n nvmet: don\u0027t check iosqes,iocqes for discovery controllers\n nvme-rdma: fix possible hang when failing to set io queues\n nvme-tcp: fix possible hang when failing to set io queues\n nvme-tcp: fix misuse of __smp_processor_id with preemption enabled\n nvme-tcp: fix a NULL deref when receiving a 0-length r2t PDU\n nvme: fix Write Zeroes limitations\n nvme: allocate the keep alive request using BLK_MQ_REQ_NOWAIT\n nvme: merge nvme_keep_alive into nvme_keep_alive_work\n nvme-fabrics: only reserve a single tag\n" }, { "commit": "38c93587375053c5b9ef093f4a5ea754538cba32", "tree": "8c6554263f39529cfb04901f006415d4fdb83cac", "parents": [ "698bacefe993ad2922c9d3b1380591ad489355e9" ], "author": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Thu Mar 18 11:31:51 2021 +0100" }, "committer": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Fri Mar 19 13:16:44 2021 +0100" }, "message": "static_call: Fix static_call_update() sanity check\n\nSites that match init_section_contains() get marked as INIT. For\nbuilt-in code init_sections contains both __init and __exit text. OTOH\nkernel_text_address() only explicitly includes __init text (and there\nare no __exit text markers).\n\nMatch what jump_label already does and ignore the warning for INIT\nsites. Also see the excellent changelog for commit: 8f35eaa5f2de\n(\"jump_label: Don\u0027t warn on __exit jump entries\")\n\nFixes: 9183c3f9ed710 (\"static_call: Add inline static call infrastructure\")\nReported-by: Sumit Garg \u003csumit.garg@linaro.org\u003e\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nAcked-by: Jarkko Sakkinen \u003cjarkko@kernel.org\u003e\nTested-by: Sumit Garg \u003csumit.garg@linaro.org\u003e\nLink: https://lkml.kernel.org/r/20210318113610.739542434@infradead.org\n" }, { "commit": "698bacefe993ad2922c9d3b1380591ad489355e9", "tree": "db3411741a396904c002645ecbcfce546a29e4f4", "parents": [ "68b1eddd421d2b16c6655eceb48918a1e896bbbc" ], "author": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Thu Mar 18 11:29:56 2021 +0100" }, "committer": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Fri Mar 19 13:16:44 2021 +0100" }, "message": "static_call: Align static_call_is_init() patching condition\n\nThe intent is to avoid writing init code after init (because the text\nmight have been freed). The code is needlessly different between\njump_label and static_call and not obviously correct.\n\nThe existing code relies on the fact that the module loader clears the\ninit layout, such that within_module_init() always fails, while\njump_label relies on the module state which is more obvious and\nmatches the kernel logic.\n\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nAcked-by: Jarkko Sakkinen \u003cjarkko@kernel.org\u003e\nTested-by: Sumit Garg \u003csumit.garg@linaro.org\u003e\nLink: https://lkml.kernel.org/r/20210318113610.636651340@infradead.org\n" }, { "commit": "68b1eddd421d2b16c6655eceb48918a1e896bbbc", "tree": "e73dd0ae1617d8563010024c710b23d979daeec0", "parents": [ "bee645788e07eea63055d261d2884ea45c2ba857" ], "author": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Thu Mar 18 11:27:19 2021 +0100" }, "committer": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Fri Mar 19 13:16:44 2021 +0100" }, "message": "static_call: Fix static_call_set_init()\n\nIt turns out that static_call_set_init() does not preserve the other\nflags; IOW. it clears TAIL if it was set.\n\nFixes: 9183c3f9ed710 (\"static_call: Add inline static call infrastructure\")\nReported-by: Sumit Garg \u003csumit.garg@linaro.org\u003e\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nAcked-by: Jarkko Sakkinen \u003cjarkko@kernel.org\u003e\nTested-by: Sumit Garg \u003csumit.garg@linaro.org\u003e\nLink: https://lkml.kernel.org/r/20210318113610.519406371@infradead.org\n" }, { "commit": "a501b048a95b79e1e34f03cac3c87ff1e9f229ad", "tree": "fa5483eade48a7996f1455e40921120b2d464139", "parents": [ "b2e9df850c58c2b36e915e7d3bed3f6107cccba6" ], "author": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Thu Mar 18 20:26:47 2021 +0100" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Fri Mar 19 12:43:41 2021 +0100" }, "message": "x86/ioapic: Ignore IRQ2 again\n\nVitaly ran into an issue with hotplugging CPU0 on an Amazon instance where\nthe matrix allocator claimed to be out of vectors. He analyzed it down to\nthe point that IRQ2, the PIC cascade interrupt, which is supposed to be not\never routed to the IO/APIC ended up having an interrupt vector assigned\nwhich got moved during unplug of CPU0.\n\nThe underlying issue is that IRQ2 for various reasons (see commit\naf174783b925 (\"x86: I/O APIC: Never configure IRQ2\" for details) is treated\nas a reserved system vector by the vector core code and is not accounted as\na regular vector. The Amazon BIOS has an routing entry of pin2 to IRQ2\nwhich causes the IO/APIC setup to claim that interrupt which is granted by\nthe vector domain because there is no sanity check. As a consequence the\nallocation counter of CPU0 underflows which causes a subsequent unplug to\nfail with:\n\n [ ... ] CPU 0 has 4294967295 vectors, 589 available. Cannot disable CPU\n\nThere is another sanity check missing in the matrix allocator, but the\nunderlying root cause is that the IO/APIC code lost the IRQ2 ignore logic\nduring the conversion to irqdomains.\n\nFor almost 6 years nobody complained about this wreckage, which might\nindicate that this requirement could be lifted, but for any system which\nactually has a PIC IRQ2 is unusable by design so any routing entry has no\neffect and the interrupt cannot be connected to a device anyway.\n\nDue to that and due to history biased paranoia reasons restore the IRQ2\nignore logic and treat it as non existent despite a routing entry claiming\notherwise.\n\nFixes: d32932d02e18 (\"x86/irq: Convert IOAPIC to use hierarchical irqdomain interfaces\")\nReported-by: Vitaly Kuznetsov \u003cvkuznets@redhat.com\u003e\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nTested-by: Vitaly Kuznetsov \u003cvkuznets@redhat.com\u003e\nCc: stable@vger.kernel.org\nLink: https://lore.kernel.org/r/20210318192819.636943062@linutronix.de\n\n\n" }, { "commit": "9ce3746d64132a561bceab6421715e7c04e85074", "tree": "e43422876fa8384356ae2de098e844b73b54cfaf", "parents": [ "f4e61f0c9add3b00bd5f2df3c814d688849b8707" ], "author": { "name": "Emanuele Giuseppe Esposito", "email": "eesposit@redhat.com", "time": "Fri Mar 19 10:16:50 2021 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Fri Mar 19 05:31:32 2021 -0400" }, "message": "documentation/kvm: additional explanations on KVM_SET_BOOT_CPU_ID\n\nThe ioctl KVM_SET_BOOT_CPU_ID fails when called after vcpu creation.\nAdd this explanation in the documentation.\n\nSigned-off-by: Emanuele Giuseppe Esposito \u003ceesposit@redhat.com\u003e\nMessage-Id: \u003c20210319091650.11967-1-eesposit@redhat.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "fb98cc0b3af2ba4d87301dff2b381b12eee35d7d", "tree": "fe3f4a920ad0314863b5d42169170eb576c80820", "parents": [ "9ceee7d0841a8f7d7644021ba7d4cc1fbc7966e3" ], "author": { "name": "Ard Biesheuvel", "email": "ardb@kernel.org", "time": "Wed Mar 10 08:33:19 2021 +0100" }, "committer": { "name": "Ard Biesheuvel", "email": "ardb@kernel.org", "time": "Fri Mar 19 07:44:28 2021 +0100" }, "message": "efi: use 32-bit alignment for efi_guid_t literals\n\nCommit 494c704f9af0 (\"efi: Use 32-bit alignment for efi_guid_t\") updated\nthe type definition of efi_guid_t to ensure that it always appears\nsufficiently aligned (the UEFI spec is ambiguous about this, but given\nthe fact that its EFI_GUID type is defined in terms of a struct carrying\na uint32_t, the natural alignment is definitely \u003e\u003d 32 bits).\n\nHowever, we missed the EFI_GUID() macro which is used to instantiate\nefi_guid_t literals: that macro is still based on the guid_t type,\nwhich does not have a minimum alignment at all. This results in warnings\nsuch as\n\n In file included from drivers/firmware/efi/mokvar-table.c:35:\n include/linux/efi.h:1093:34: warning: passing 1-byte aligned argument to\n 4-byte aligned parameter 2 of \u0027get_var\u0027 may result in an unaligned pointer\n access [-Walign-mismatch]\n status \u003d get_var(L\"SecureBoot\", \u0026EFI_GLOBAL_VARIABLE_GUID, NULL, \u0026size,\n ^\n include/linux/efi.h:1101:24: warning: passing 1-byte aligned argument to\n 4-byte aligned parameter 2 of \u0027get_var\u0027 may result in an unaligned pointer\n access [-Walign-mismatch]\n get_var(L\"SetupMode\", \u0026EFI_GLOBAL_VARIABLE_GUID, NULL, \u0026size, \u0026setupmode);\n\nThe distinction only matters on CPUs that do not support misaligned loads\nfully, but 32-bit ARM\u0027s load-multiple instructions fall into that category,\nand these are likely to be emitted by the compiler that built the firmware\nfor loading word-aligned 128-bit GUIDs from memory\n\nSo re-implement the initializer in terms of our own efi_guid_t type, so that\nthe alignment becomes a property of the literal\u0027s type.\n\nFixes: 494c704f9af0 (\"efi: Use 32-bit alignment for efi_guid_t\")\nReported-by: Nathan Chancellor \u003cnathan@kernel.org\u003e\nReviewed-by: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nReviewed-by: Nathan Chancellor \u003cnathan@kernel.org\u003e\nTested-by: Nathan Chancellor \u003cnathan@kernel.org\u003e\nLink: https://github.com/ClangBuiltLinux/linux/issues/1327\nSigned-off-by: Ard Biesheuvel \u003cardb@kernel.org\u003e\n" }, { "commit": "9ceee7d0841a8f7d7644021ba7d4cc1fbc7966e3", "tree": "91df4aa2a0987806b89ba4b033591dec45327bbb", "parents": [ "483028edacab374060d93955382b4865a9e07cba" ], "author": { "name": "Lv Yunlong", "email": "lyl2019@mail.ustc.edu.cn", "time": "Wed Mar 10 00:31:27 2021 -0800" }, "committer": { "name": "Ard Biesheuvel", "email": "ardb@kernel.org", "time": "Fri Mar 19 07:44:27 2021 +0100" }, "message": "firmware/efi: Fix a use after bug in efi_mem_reserve_persistent\n\nIn the for loop in efi_mem_reserve_persistent(), prsv \u003d rsv-\u003enext\nuse the unmapped rsv. Use the unmapped pages will cause segment\nfault.\n\nFixes: 18df7577adae6 (\"efi/memreserve: deal with memreserve entries in unmapped memory\")\nSigned-off-by: Lv Yunlong \u003clyl2019@mail.ustc.edu.cn\u003e\nSigned-off-by: Ard Biesheuvel \u003cardb@kernel.org\u003e\n" }, { "commit": "af3ef3b1031634724a3763606695ebcd113d782b", "tree": "589e504c57af5b35526dac8c8e84d33a633eea82", "parents": [ "403dba003d17b3f0c1627b355cec2d74041cf648" ], "author": { "name": "Aurelien Aptel", "email": "aaptel@suse.com", "time": "Thu Mar 18 19:17:10 2021 +0100" }, "committer": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Fri Mar 19 00:50:58 2021 -0500" }, "message": "cifs: warn and fail if trying to use rootfs without the config option\n\nIf CONFIG_CIFS_ROOT is not set, rootfs mount option is invalid\n\nSigned-off-by: Aurelien Aptel \u003caaptel@suse.com\u003e\nCC: \u003cstable@vger.kernel.org\u003e # v5.11\nSigned-off-by: Steve French \u003cstfrench@microsoft.com\u003e\n" }, { "commit": "403dba003d17b3f0c1627b355cec2d74041cf648", "tree": "1ff9ad6bce3b9cb358ffffa26b86f4057d88a542", "parents": [ "05946d4b7a7349ae58bfa2d51ae832e64a394c2d" ], "author": { "name": "Liu xuzhi", "email": "liu.xuzhi@zte.com.cn", "time": "Thu Mar 18 17:46:57 2021 -0700" }, "committer": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Fri Mar 19 00:37:51 2021 -0500" }, "message": "fs/cifs/: fix misspellings using codespell tool\n\nA typo is found out by codespell tool in 251th lines of cifs_swn.c:\n\n$ codespell ./fs/cifs/\n./cifs_swn.c:251: funciton \u003d\u003d\u003e function\n\nFix a typo found by codespell.\n\nSigned-off-by: Liu xuzhi \u003cliu.xuzhi@zte.com.cn\u003e\nSigned-off-by: Steve French \u003cstfrench@microsoft.com\u003e\n" }, { "commit": "ee5de60a08b7d8d255722662da461ea159c15538", "tree": "8838fb5084cd50ce46875f44e030807c13310f84", "parents": [ "6406887a12ee5dcdaffff1a8508d91113d545559" ], "author": { "name": "Ondrej Mosnacek", "email": "omosnace@redhat.com", "time": "Thu Mar 18 22:53:03 2021 +0100" }, "committer": { "name": "Paul Moore", "email": "paul@paul-moore.com", "time": "Thu Mar 18 23:26:59 2021 -0400" }, "message": "selinuxfs: unify policy load error reporting\n\nLet\u0027s drop the pr_err()s from sel_make_policy_nodes() and just add one\npr_warn_ratelimited() call to the sel_make_policy_nodes() error path in\nsel_write_load().\n\nChanging from error to warning makes sense, since after 02a52c5c8c3b\n(\"selinux: move policy commit after updating selinuxfs\"), this error\npath no longer leads to a broken selinuxfs tree (it\u0027s just kept in the\noriginal state and policy load is aborted).\n\nI also added _ratelimited to be consistent with the other prtin in the\nsame function (it\u0027s probably not necessary, but can\u0027t really hurt...\nthere are likely more important error messages to be printed when\nfilesystem entry creation starts erroring out).\n\nSuggested-by: Paul Moore \u003cpaul@paul-moore.com\u003e\nSigned-off-by: Ondrej Mosnacek \u003comosnace@redhat.com\u003e\nSigned-off-by: Paul Moore \u003cpaul@paul-moore.com\u003e\n" }, { "commit": "6406887a12ee5dcdaffff1a8508d91113d545559", "tree": "330e72853653a1f8071af6e1cc304ae3ceae304e", "parents": [ "519dad3bcd809dc1523bf80ab0310ddb3bf00ade" ], "author": { "name": "Ondrej Mosnacek", "email": "omosnace@redhat.com", "time": "Thu Mar 18 22:53:02 2021 +0100" }, "committer": { "name": "Paul Moore", "email": "paul@paul-moore.com", "time": "Thu Mar 18 23:23:46 2021 -0400" }, "message": "selinux: fix variable scope issue in live sidtab conversion\n\nCommit 02a52c5c8c3b (\"selinux: move policy commit after updating\nselinuxfs\") moved the selinux_policy_commit() call out of\nsecurity_load_policy() into sel_write_load(), which caused a subtle yet\nrather serious bug.\n\nThe problem is that security_load_policy() passes a reference to the\nconvert_params local variable to sidtab_convert(), which stores it in\nthe sidtab, where it may be accessed until the policy is swapped over\nand RCU synchronized. Before 02a52c5c8c3b, selinux_policy_commit() was\ncalled directly from security_load_policy(), so the convert_params\npointer remained valid all the way until the old sidtab was destroyed,\nbut now that\u0027s no longer the case and calls to sidtab_context_to_sid()\non the old sidtab after security_load_policy() returns may cause invalid\nmemory accesses.\n\nThis can be easily triggered using the stress test from commit\nee1a84fdfeed (\"selinux: overhaul sidtab to fix bug and improve\nperformance\"):\n```\nfunction rand_cat() {\n\techo $(( $RANDOM % 1024 ))\n}\n\nfunction do_work() {\n\twhile true; do\n\t\techo -n \"system_u:system_r:kernel_t:s0:c$(rand_cat),c$(rand_cat)\" \\\n\t\t\t\u003e/sys/fs/selinux/context 2\u003e/dev/null || true\n\tdone\n}\n\ndo_work \u003e/dev/null \u0026\ndo_work \u003e/dev/null \u0026\ndo_work \u003e/dev/null \u0026\n\nwhile load_policy; do echo -n .; sleep 0.1; done\n\nkill %1\nkill %2\nkill %3\n```\n\nFix this by allocating the temporary sidtab convert structures\ndynamically and passing them among the\nselinux_policy_{load,cancel,commit} functions.\n\nFixes: 02a52c5c8c3b (\"selinux: move policy commit after updating selinuxfs\")\nCc: stable@vger.kernel.org\nTested-by: Tyler Hicks \u003ctyhicks@linux.microsoft.com\u003e\nReviewed-by: Tyler Hicks \u003ctyhicks@linux.microsoft.com\u003e\nSigned-off-by: Ondrej Mosnacek \u003comosnace@redhat.com\u003e\n[PM: merge fuzz in security.h and services.c]\nSigned-off-by: Paul Moore \u003cpaul@paul-moore.com\u003e\n" }, { "commit": "519dad3bcd809dc1523bf80ab0310ddb3bf00ade", "tree": "b01c0dacf941d602dbdce10a105e4bba14ee3949", "parents": [ "365982aba1f264dba26f0908700d62bfa046918c" ], "author": { "name": "Ondrej Mosnacek", "email": "omosnace@redhat.com", "time": "Thu Mar 18 22:53:01 2021 +0100" }, "committer": { "name": "Paul Moore", "email": "paul@paul-moore.com", "time": "Thu Mar 18 23:13:04 2021 -0400" }, "message": "selinux: don\u0027t log MAC_POLICY_LOAD record on failed policy load\n\nIf sel_make_policy_nodes() fails, we should jump to \u0027out\u0027, not \u0027out1\u0027,\nas the latter would incorrectly log an MAC_POLICY_LOAD audit record,\neven though the policy hasn\u0027t actually been reloaded. The \u0027out1\u0027 jump\nlabel now becomes unused and can be removed.\n\nFixes: 02a52c5c8c3b (\"selinux: move policy commit after updating selinuxfs\")\nCc: stable@vger.kernel.org\nSigned-off-by: Ondrej Mosnacek \u003comosnace@redhat.com\u003e\nSigned-off-by: Paul Moore \u003cpaul@paul-moore.com\u003e\n" }, { "commit": "8b12a62a4e3ed4ae99c715034f557eb391d6b196", "tree": "14d93a9750ed512e36ef16a61e8b3796d5cb442a", "parents": [ "81aa0968b7ea6dbabcdcda37dc8434dca6e1565b", "e94c55b8e0a0bbe9a026250cf31e2fa45957d776" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 17:57:34 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 17:57:34 2021 -0700" }, "message": "Merge tag \u0027drm-fixes-2021-03-19\u0027 of git://anongit.freedesktop.org/drm/drm\n\nPull drm fixes from Dave Airlie:\n \"Regular fixes pull, pretty small set of fixes, a couple of i915 and\n amdgpu, one ttm, one nouveau and one omap. Probably smaller than usual\n for this time, so we\u0027ll see if something pops up next week or if this\n will continue to stay small.\n\n Summary:\n\n ttm:\n - Make ttm_bo_unpin() not wraparound on too many unpins\n\n omap:\n - Fix coccicheck warning in omap\n\n amdgpu:\n - DCN 3.0 gamma fixes\n - DCN 2.1 corrupt screen fix\n\n i915:\n - Workaround async flip + VT-d frame corruption on HSW/BDW\n - Fix NMI watchdog crash due to uninitialized OA buffer use on gen12+\n\n nouveau:\n - workaround oops with bo syncing\"\n\n* tag \u0027drm-fixes-2021-03-19\u0027 of git://anongit.freedesktop.org/drm/drm:\n nouveau: Skip unvailable ttm page entries\n drm/amd/display: Remove MPC gamut remap logic for DCN30\n drm/amd/display: Correct algorithm for reversed gamma\n drm/omap: dsi: fix unsigned expression compared with zero\n i915/perf: Start hrtimer only if sampling the OA buffer\n drm/i915: Workaround async flip + VT-d corruption on HSW/BDW\n drm/amd/display: Copy over soc values before bounding box creation\n drm/ttm: make ttm_bo_unpin more defensive\n" }, { "commit": "e94c55b8e0a0bbe9a026250cf31e2fa45957d776", "tree": "d2e6bfb7e96e0224feb652040d0fb5215769ca44", "parents": [ "0677170bcf5c3aba0cde1a7b9e405532b7441bf4" ], "author": { "name": "Tobias Klausmann", "email": "tobias.klausmann@freenet.de", "time": "Sat Mar 13 23:21:59 2021 +0100" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Mar 19 10:38:31 2021 +1000" }, "message": "nouveau: Skip unvailable ttm page entries\n\nStarting with commit f295c8cfec833c2707ff1512da10d65386dde7af\n(\"drm/nouveau: fix dma syncing warning with debugging on.\")\nthe following oops occures:\n\n BUG: kernel NULL pointer dereference, address: 0000000000000000\n #PF: supervisor read access in kernel mode\n #PF: error_code(0x0000) - not-present page\n PGD 0 P4D 0\n Oops: 0000 [#1] PREEMPT SMP PTI\n CPU: 6 PID: 1013 Comm: Xorg.bin Tainted: G E 5.11.0-desktop-rc0+ #2\n Hardware name: Acer Aspire VN7-593G/Pluto_KLS, BIOS V1.11 08/01/2018\n RIP: 0010:nouveau_bo_sync_for_device+0x40/0xb0 [nouveau]\n Call Trace:\n nouveau_bo_validate+0x5d/0x80 [nouveau]\n nouveau_gem_ioctl_pushbuf+0x662/0x1120 [nouveau]\n ? nouveau_gem_ioctl_new+0xf0/0xf0 [nouveau]\n drm_ioctl_kernel+0xa6/0xf0 [drm]\n drm_ioctl+0x1f4/0x3a0 [drm]\n ? nouveau_gem_ioctl_new+0xf0/0xf0 [nouveau]\n nouveau_drm_ioctl+0x50/0xa0 [nouveau]\n __x64_sys_ioctl+0x7e/0xb0\n do_syscall_64+0x33/0x80\n entry_SYSCALL_64_after_hwframe+0x44/0xae\n ---[ end trace ccfb1e7f4064374f ]---\n RIP: 0010:nouveau_bo_sync_for_device+0x40/0xb0 [nouveau]\n\nThe underlying problem is not introduced by the commit, yet it uncovered the\nunderlying issue. The cited commit relies on valid pages. This is not given for\ndue to some bugs. For now, just warn and work around the issue by just ignoring\nthe bad ttm objects.\nBelow is some debug info gathered while debugging this issue:\n\nnouveau 0000:01:00.0: DRM: ttm_dma-\u003enum_pages: 2048\nnouveau 0000:01:00.0: DRM: ttm_dma-\u003epages is NULL\nnouveau 0000:01:00.0: DRM: ttm_dma: 00000000e96058e7\nnouveau 0000:01:00.0: DRM: ttm_dma-\u003epage_flags:\nnouveau 0000:01:00.0: DRM: ttm_dma: Populated: 1\nnouveau 0000:01:00.0: DRM: ttm_dma: No Retry: 0\nnouveau 0000:01:00.0: DRM: ttm_dma: SG: 256\nnouveau 0000:01:00.0: DRM: ttm_dma: Zero Alloc: 0\nnouveau 0000:01:00.0: DRM: ttm_dma: Swapped: 0\n\nSigned-off-by: Tobias Klausmann \u003ctobias.klausmann@freenet.de\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20210313222159.3346-1-tobias.klausmann@freenet.de\n" }, { "commit": "0677170bcf5c3aba0cde1a7b9e405532b7441bf4", "tree": "9b54725f50d375e0716babe881523fe9e8ae56ec", "parents": [ "a97fdabcf2b4ff78b03951529dc77469f3646fd3", "6a77c6bb7260bd5000f95df454d9f8cdb1af7132" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Mar 19 10:33:32 2021 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Mar 19 10:33:32 2021 +1000" }, "message": "Merge tag \u0027drm-intel-fixes-2021-03-18\u0027 of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes\n\ndrm/i915 fixes for v5.12-rc4:\n- Workaround async flip + VT-d frame corruption on HSW/BDW\n- Fix NMI watchdog crash due to uninitialized OA buffer use on gen12+\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nFrom: Jani Nikula \u003cjani.nikula@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/87blbg8y5t.fsf@intel.com\n" }, { "commit": "a97fdabcf2b4ff78b03951529dc77469f3646fd3", "tree": "626d9a61250755ed4f46abaed4f399e2503c510d", "parents": [ "db3f0d8ee9bf4b9fb7f9c8bbea3e5fad0cd9b66e", "beb6b2f97e0a02164c7f0df6e08c49219cfc2b80" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Mar 19 10:31:57 2021 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Mar 19 10:31:58 2021 +1000" }, "message": "Merge tag \u0027amd-drm-fixes-5.12-2021-03-18\u0027 of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes\n\namdgpu:\n- DCN 3.0 gamma fixes\n- DCN 2.1 corrupt screen fix\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nFrom: Alex Deucher \u003calexander.deucher@amd.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20210318042858.3810-1-alexander.deucher@amd.com\n" }, { "commit": "db3f0d8ee9bf4b9fb7f9c8bbea3e5fad0cd9b66e", "tree": "f00a9dae3ebd557e891f695bf985f452c34d1fb4", "parents": [ "1e28eed17697bcf343c6743f0028cc3b5dd88bf0", "6909115442759efef3d4bc5d9c54d7943f1afc14" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Mar 19 10:21:33 2021 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Mar 19 10:21:48 2021 +1000" }, "message": "Merge tag \u0027drm-misc-fixes-2021-03-18\u0027 of git://anongit.freedesktop.org/drm/drm-misc into drm-fixes\n\ndrm-misc-fixes for v5.12-rc4:\n- Make ttm_bo_unpin() not wraparound on too many unpins.\n- Fix coccicheck warning in omap.\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Maarten Lankhorst \u003cmaarten.lankhorst@linux.intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/a0e13bbb-6ba6-ff24-4db8-0e02e605de18@linux.intel.com\n" }, { "commit": "81aa0968b7ea6dbabcdcda37dc8434dca6e1565b", "tree": "9cbc9fc23517c3dd6b6f89c0b0557edbb438b8b3", "parents": [ "dc0337999d87a5e749ef1ac0bcc1a06d2a3f9ec0", "485df75554257e883d0ce39bb886e8212349748e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 13:38:42 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 13:38:42 2021 -0700" }, "message": "Merge tag \u0027for-5.12-rc3-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux\n\nPull btrfs fixes from David Sterba:\n \"There are still regressions being found and fixed in the zoned mode\n and subpage code, the rest are fixes for bugs reported by users.\n\n Regressions:\n\n - subpage block support:\n - readahead works on the proper block size\n - fix last page zeroing\n\n - zoned mode:\n - linked list corruption for tree log\n\n Fixes:\n\n - qgroup leak after falloc failure\n\n - tree mod log and backref resolving:\n - extent buffer cloning race when resolving backrefs\n - pin deleted leaves with active tree mod log users\n\n - drop debugging flag from slab cache\"\n\n* tag \u0027for-5.12-rc3-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:\n btrfs: always pin deleted leaves when there are active tree mod log users\n btrfs: fix race when cloning extent buffer during rewind of an old root\n btrfs: fix slab cache flags for free space tree bitmap\n btrfs: subpage: make readahead work properly\n btrfs: subpage: fix wild pointer access during metadata read failure\n btrfs: zoned: fix linked list corruption after log root tree allocation failure\n btrfs: fix qgroup data rsv leak caused by falloc failure\n btrfs: track qgroup released data in own variable in insert_prealloc_file_extent\n btrfs: fix wrong offset to zero out range beyond i_size\n" }, { "commit": "dc0337999d87a5e749ef1ac0bcc1a06d2a3f9ec0", "tree": "558f7fcfd5308e4ad69555d196225c4f8a329a47", "parents": [ "c73891c922f5934b826fe5eb743fbdb28aee3f99", "4ab4fcfce5b540227d80eb32f1db45ab615f7c92" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 12:37:05 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 12:37:05 2021 -0700" }, "message": "Merge tag \u0027vfio-v5.12-rc4\u0027 of git://github.com/awilliam/linux-vfio\n\nPull VFIO fixes from Alex Williamson:\n\n - Fix 32-bit issue with new unmap-all flag (Steve Sistare)\n\n - Various Kconfig changes for better coverage (Jason Gunthorpe)\n\n - Fix to batch pinning support (Daniel Jordan)\n\n* tag \u0027vfio-v5.12-rc4\u0027 of git://github.com/awilliam/linux-vfio:\n vfio/type1: fix vaddr_get_pfns() return in vfio_pin_page_external()\n vfio: Depend on MMU\n ARM: amba: Allow some ARM_AMBA users to compile with COMPILE_TEST\n vfio-platform: Add COMPILE_TEST to VFIO_PLATFORM\n vfio: IOMMU_API should be selected\n vfio/type1: fix unmap all on ILP32\n" }, { "commit": "c73891c922f5934b826fe5eb743fbdb28aee3f99", "tree": "658d30a1a92a99ee18fc1bdd69a9647c40b4e425", "parents": [ "bf152b0b41dc141c8d32eb6e974408f5804f4d00", "8723d5ba8bdae1c41be7a6fc8469dc9aa551e7d0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 12:32:51 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 12:32:51 2021 -0700" }, "message": "Merge tag \u0027xfs-5.12-fixes-3\u0027 of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux\n\nPull xfs fixes from Darrick Wong:\n \"A couple of minor corrections for the new idmapping functionality, and\n a fix for a theoretical hang that could occur if we decide to abort a\n mount after dirtying the quota inodes.\n\n Summary:\n\n - Fix quota accounting on creat() when id mapping is enabled\n\n - Actually reclaim dirty quota inodes when mount fails\n\n - Typo fixes for documentation\n\n - Restrict both bulkstat calls on idmapped/namespaced mounts\"\n\n* tag \u0027xfs-5.12-fixes-3\u0027 of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:\n xfs: also reject BULKSTAT_SINGLE in a mount user namespace\n docs: ABI: Fix the spelling oustanding to outstanding in the file sysfs-fs-xfs\n xfs: force log and push AIL to clear pinned inodes when aborting mount\n xfs: fix quota accounting when a mount is idmapped\n" }, { "commit": "bf152b0b41dc141c8d32eb6e974408f5804f4d00", "tree": "792b4b0e31d8d9df695958c1bccaf9ff537d73d2", "parents": [ "8ff0f3bf5d6513dfb7462246d9c656da7c02b37e", "0bde59c1723a29e294765c96dbe5c7fb639c2f96" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 11:20:35 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 11:20:35 2021 -0700" }, "message": "Merge tag \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost\n\nPull virtio fixes from Michael Tsirkin:\n \"Some fixes and cleanups all over the place\"\n\n* tag \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:\n vhost-vdpa: set v-\u003econfig_ctx to NULL if eventfd_ctx_fdget() fails\n vhost-vdpa: fix use-after-free of v-\u003econfig_ctx\n vhost: Fix vhost_vq_reset()\n vhost_vdpa: fix the missing irq_bypass_unregister_producer() invocation\n vdpa_sim: Skip typecasting from void*\n virtio: remove export for virtio_config_{enable, disable}\n virtio-mmio: Use to_virtio_mmio_device() to simply code\n vdpa: set the virtqueue num during register\n" }, { "commit": "f4e61f0c9add3b00bd5f2df3c814d688849b8707", "tree": "288941d008a8996551da972d7db2bcf283961e78", "parents": [ "c2162e13d6e2f43e5001a356196871642de070ba" ], "author": { "name": "Wanpeng Li", "email": "wanpengli@tencent.com", "time": "Mon Mar 15 14:55:28 2021 +0800" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 18 13:58:14 2021 -0400" }, "message": "x86/kvm: Fix broken irq restoration in kvm_wait\n\nAfter commit 997acaf6b4b59c (lockdep: report broken irq restoration), the guest\nsplatting below during boot:\n\n raw_local_irq_restore() called with IRQs enabled\n WARNING: CPU: 1 PID: 169 at kernel/locking/irqflag-debug.c:10 warn_bogus_irq_restore+0x26/0x30\n Modules linked in: hid_generic usbhid hid\n CPU: 1 PID: 169 Comm: systemd-udevd Not tainted 5.11.0+ #25\n RIP: 0010:warn_bogus_irq_restore+0x26/0x30\n Call Trace:\n kvm_wait+0x76/0x90\n __pv_queued_spin_lock_slowpath+0x285/0x2e0\n do_raw_spin_lock+0xc9/0xd0\n _raw_spin_lock+0x59/0x70\n lockref_get_not_dead+0xf/0x50\n __legitimize_path+0x31/0x60\n legitimize_root+0x37/0x50\n try_to_unlazy_next+0x7f/0x1d0\n lookup_fast+0xb0/0x170\n path_openat+0x165/0x9b0\n do_filp_open+0x99/0x110\n do_sys_openat2+0x1f1/0x2e0\n do_sys_open+0x5c/0x80\n __x64_sys_open+0x21/0x30\n do_syscall_64+0x32/0x50\n entry_SYSCALL_64_after_hwframe+0x44/0xae\n\nThe new consistency checking, expects local_irq_save() and\nlocal_irq_restore() to be paired and sanely nested, and therefore expects\nlocal_irq_restore() to be called with irqs disabled.\nThe irqflags handling in kvm_wait() which ends up doing:\n\n\tlocal_irq_save(flags);\n\tsafe_halt();\n\tlocal_irq_restore(flags);\n\ninstead triggers it. This patch fixes it by using\nlocal_irq_disable()/enable() directly.\n\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nReported-by: Dmitry Vyukov \u003cdvyukov@google.com\u003e\nSigned-off-by: Wanpeng Li \u003cwanpengli@tencent.com\u003e\nMessage-Id: \u003c1615791328-2735-1-git-send-email-wanpengli@tencent.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "c2162e13d6e2f43e5001a356196871642de070ba", "tree": "02e4a1d84b6df817d1b5c91f69d2178e5da9980e", "parents": [ "b318e8decf6b9ef1bcf4ca06fae6d6a2cb5d5c5c" ], "author": { "name": "Wanpeng Li", "email": "wanpengli@tencent.com", "time": "Fri Mar 12 10:45:51 2021 +0800" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 18 13:55:34 2021 -0400" }, "message": "KVM: X86: Fix missing local pCPU when executing wbinvd on all dirty pCPUs\n\nIn order to deal with noncoherent DMA, we should execute wbinvd on\nall dirty pCPUs when guest wbinvd exits to maintain data consistency.\nsmp_call_function_many() does not execute the provided function on the\nlocal core, therefore replace it by on_each_cpu_mask().\n\nReported-by: Nadav Amit \u003cnamit@vmware.com\u003e\nCc: Nadav Amit \u003cnamit@vmware.com\u003e\nSigned-off-by: Wanpeng Li \u003cwanpengli@tencent.com\u003e\nMessage-Id: \u003c1615517151-7465-1-git-send-email-wanpengli@tencent.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "b318e8decf6b9ef1bcf4ca06fae6d6a2cb5d5c5c", "tree": "3904104c1dbef1188efa8c45ff6617c658c0b271", "parents": [ "3df2252436c08028a549e27ed7f097974e21d17b" ], "author": { "name": "Sean Christopherson", "email": "seanjc@google.com", "time": "Tue Mar 16 11:44:33 2021 -0700" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 18 13:55:14 2021 -0400" }, "message": "KVM: x86: Protect userspace MSR filter with SRCU, and set atomically-ish\n\nFix a plethora of issues with MSR filtering by installing the resulting\nfilter as an atomic bundle instead of updating the live filter one range\nat a time. The KVM_X86_SET_MSR_FILTER ioctl() isn\u0027t truly atomic, as\nthe hardware MSR bitmaps won\u0027t be updated until the next VM-Enter, but\nthe relevant software struct is atomically updated, which is what KVM\nreally needs.\n\nSimilar to the approach used for modifying memslots, make arch.msr_filter\na SRCU-protected pointer, do all the work configuring the new filter\noutside of kvm-\u003elock, and then acquire kvm-\u003elock only when the new filter\nhas been vetted and created. That way vCPU readers either see the old\nfilter or the new filter in their entirety, not some half-baked state.\n\nYuan Yao pointed out a use-after-free in ksm_msr_allowed() due to a\nTOCTOU bug, but that\u0027s just the tip of the iceberg...\n\n - Nothing is __rcu annotated, making it nigh impossible to audit the\n code for correctness.\n - kvm_add_msr_filter() has an unpaired smp_wmb(). Violation of kernel\n coding style aside, the lack of a smb_rmb() anywhere casts all code\n into doubt.\n - kvm_clear_msr_filter() has a double free TOCTOU bug, as it grabs\n count before taking the lock.\n - kvm_clear_msr_filter() also has memory leak due to the same TOCTOU bug.\n\nThe entire approach of updating the live filter is also flawed. While\ninstalling a new filter is inherently racy if vCPUs are running, fixing\nthe above issues also makes it trivial to ensure certain behavior is\ndeterministic, e.g. KVM can provide deterministic behavior for MSRs with\nidentical settings in the old and new filters. An atomic update of the\nfilter also prevents KVM from getting into a half-baked state, e.g. if\ninstalling a filter fails, the existing approach would leave the filter\nin a half-baked state, having already committed whatever bits of the\nfilter were already processed.\n\n[*] https://lkml.kernel.org/r/20210312083157.25403-1-yaoyuan0329os@gmail.com\n\nFixes: 1a155254ff93 (\"KVM: x86: Introduce MSR filtering\")\nCc: stable@vger.kernel.org\nCc: Alexander Graf \u003cgraf@amazon.com\u003e\nReported-by: Yuan Yao \u003cyaoyuan0329os@gmail.com\u003e\nSigned-off-by: Sean Christopherson \u003cseanjc@google.com\u003e\nMessage-Id: \u003c20210316184436.2544875-2-seanjc@google.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "3df2252436c08028a549e27ed7f097974e21d17b", "tree": "dd65e8920ab2fd9b09f0a9eebac23d5eec1c91d9", "parents": [ "e2c12909ae5f5181d9e0b0c536e26c6877daec48" ], "author": { "name": "Emanuele Giuseppe Esposito", "email": "eesposit@redhat.com", "time": "Thu Mar 18 16:16:24 2021 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 18 13:55:14 2021 -0400" }, "message": "selftests: kvm: add set_boot_cpu_id test\n\nTest for the KVM_SET_BOOT_CPU_ID ioctl.\nCheck that it correctly allows to change the BSP vcpu.\n\nSigned-off-by: Emanuele Giuseppe Esposito \u003ceesposit@redhat.com\u003e\nMessage-Id: \u003c20210318151624.490861-2-eesposit@redhat.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "e2c12909ae5f5181d9e0b0c536e26c6877daec48", "tree": "2622cc07ccdec558dbfe02efb95e96d019083f3e", "parents": [ "77a3aa26a00fe55325ae2a51d80a56836d1edce8" ], "author": { "name": "Emanuele Giuseppe Esposito", "email": "eesposit@redhat.com", "time": "Thu Mar 18 16:16:23 2021 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 18 13:55:14 2021 -0400" }, "message": "selftests: kvm: add _vm_ioctl\n\nAs in kvm_ioctl and _kvm_ioctl, add\nthe respective _vm_ioctl for vm_ioctl.\n\n_vm_ioctl invokes an ioctl using the vm fd,\nleaving the caller to test the result.\n\nSigned-off-by: Emanuele Giuseppe Esposito \u003ceesposit@redhat.com\u003e\nMessage-Id: \u003c20210318151624.490861-1-eesposit@redhat.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "8ff0f3bf5d6513dfb7462246d9c656da7c02b37e", "tree": "03e889ee36115d505b2dc822a907f5e4cbed1b91", "parents": [ "6417f03132a6952cd17ddd8eaddbac92b61b17e0", "5808fecc572391867fcd929662b29c12e6d08d81" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 10:37:30 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 18 10:37:30 2021 -0700" }, "message": "Merge branch \u0027iomap-5.12-fixes\u0027 of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux\n\nPull iomap fix from Darrick Wong:\n \"A single fix to the iomap code which fixes some drama when someone\n gives us a {de,ma}liciously fragmented swap file\"\n\n* \u0027iomap-5.12-fixes\u0027 of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:\n iomap: Fix negative assignment to unsigned sis-\u003epages in iomap_swapfile_activate\n" }, { "commit": "77a3aa26a00fe55325ae2a51d80a56836d1edce8", "tree": "3148ce74decd74d0da367f857c6c6282524190f2", "parents": [ "2c7f76b4c42bd5d953bc821e151644434865f999" ], "author": { "name": "Emanuele Giuseppe Esposito", "email": "eesposit@redhat.com", "time": "Thu Mar 18 15:56:29 2021 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 18 13:30:49 2021 -0400" }, "message": "selftests: kvm: add get_msr_index_features\n\nTest the KVM_GET_MSR_FEATURE_INDEX_LIST\nand KVM_GET_MSR_INDEX_LIST ioctls.\n\nSigned-off-by: Emanuele Giuseppe Esposito \u003ceesposit@redhat.com\u003e\nMessage-Id: \u003c20210318145629.486450-1-eesposit@redhat.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "83b62687a05205847d627f29126a8fee3c644335", "tree": "cc9c37a29f9de97cfd17e4fd9dfc8663ebc72b4f", "parents": [ "1e28eed17697bcf343c6743f0028cc3b5dd88bf0" ], "author": { "name": "Steven Rostedt (VMware)", "email": "rostedt@goodmis.org", "time": "Thu Mar 18 11:44:08 2021 -0400" }, "committer": { "name": "Steven Rostedt (VMware)", "email": "rostedt@goodmis.org", "time": "Thu Mar 18 12:57:37 2021 -0400" }, "message": "workqueue/tracing: Copy workqueue name to buffer in trace event\n\nThe trace event \"workqueue_queue_work\" references an unsafe string in\ndereferencing the name of the workqueue. As the name is allocated, it\ncould later be freed, and the pointer to that string could stay on the\ntracing buffer. If the trace buffer is read after the string is freed, it\nwill reference an unsafe pointer.\n\nI added a new verifier to make sure that all strings referenced in the\noutput of the trace buffer is safe to read and this triggered on the\nworkqueue_queue_work trace event:\n\nworkqueue_queue_work: work struct\u003d00000000b2b235c7 function\u003dgc_worker workqueue\u003d(0xffff888100051160:events_power_efficient)[UNSAFE-MEMORY] req_cpu\u003d256 cpu\u003d1\nworkqueue_queue_work: work struct\u003d00000000c344caec function\u003dflush_to_ldisc workqueue\u003d(0xffff888100054d60:events_unbound)[UNSAFE-MEMORY] req_cpu\u003d256 cpu\u003d4294967295\nworkqueue_queue_work: work struct\u003d00000000b2b235c7 function\u003dgc_worker workqueue\u003d(0xffff888100051160:events_power_efficient)[UNSAFE-MEMORY] req_cpu\u003d256 cpu\u003d1\nworkqueue_queue_work: work struct\u003d000000000b238b3f function\u003dvmstat_update workqueue\u003d(0xffff8881000c3760:mm_percpu_wq)[UNSAFE-MEMORY] req_cpu\u003d1 cpu\u003d1\n\nAlso, if this event is read via a user space application like perf or\ntrace-cmd, the name would only be an address and useless information:\n\nworkqueue_queue_work: work struct\u003d0xffff953f80b4b918 function\u003ddisk_events_workfn workqueue\u003dffff953f8005d378 req_cpu\u003d8192 cpu\u003d5\n\nCc: Zqiang \u003cqiang.zhang@windriver.com\u003e\nCc: Tejun Heo \u003ctj@kernel.org\u003e\nFixes: 7bf9c4a88e3e3 (\"workqueue: tracing the name of the workqueue instead of it\u0027s address\")\nSigned-off-by: Steven Rostedt (VMware) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "9d3fcb28f9b9750b474811a2964ce022df56336e", "tree": "d9d49bd8eea74c47ee6aa2f16e314c839714927d", "parents": [ "1e28eed17697bcf343c6743f0028cc3b5dd88bf0" ], "author": { "name": "Josef Bacik", "email": "josef@toxicpanda.com", "time": "Tue Mar 16 22:17:48 2021 -0400" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Mar 18 16:58:02 2021 +0100" }, "message": "Revert \"PM: ACPI: reboot: Use S5 for reboot\"\n\nThis reverts commit d60cd06331a3566d3305b3c7b566e79edf4e2095.\n\nThis patch causes a panic when rebooting my Dell Poweredge r440. I do\nnot have the full panic log as it\u0027s lost at that stage of the reboot and\nI do not have a serial console. Reverting this patch makes my system\nable to reboot again.\n\nSigned-off-by: Josef Bacik \u003cjosef@toxicpanda.com\u003e\nSigned-off-by: Rafael J. Wysocki \u003crafael.j.wysocki@intel.com\u003e\n" }, { "commit": "de75a3d3f5a14c9ab3c4883de3471d3c92a8ee78", "tree": "825ae49269672936357ee4e265b31db25d6ecf64", "parents": [ "ee53fb2b197b72b126ca0387ae636da75d969428" ], "author": { "name": "Pavel Begunkov", "email": "asml.silence@gmail.com", "time": "Thu Mar 18 11:54:35 2021 +0000" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Mar 18 09:44:35 2021 -0600" }, "message": "io_uring: don\u0027t leak creds on SQO attach error\n\nAttaching to already dead/dying SQPOLL task is disallowed in\nio_sq_offload_create(), but cleanup is hand coded by calling\nio_put_sq_data()/etc., that miss to put ctx-\u003esq_creds.\n\nDefer everything to error-path io_sq_thread_finish(), adding\nctx-\u003esqd_list in the error case as well as finish will handle it.\n\nSigned-off-by: Pavel Begunkov \u003casml.silence@gmail.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "ee53fb2b197b72b126ca0387ae636da75d969428", "tree": "7a96187881cdc2597ca22ebed16c1d79815e78b8", "parents": [ "53e043b2b432ef2294efec04dd8a88d96c024624" ], "author": { "name": "Stefan Metzmacher", "email": "metze@samba.org", "time": "Mon Mar 15 12:56:57 2021 +0100" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Mar 18 09:44:35 2021 -0600" }, "message": "io_uring: use typesafe pointers in io_uring_task\n\nSigned-off-by: Stefan Metzmacher \u003cmetze@samba.org\u003e\nLink: https://lore.kernel.org/r/ce2a598e66e48347bb04afbaf2acc67c0cc7971a.1615809009.git.metze@samba.org\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "53e043b2b432ef2294efec04dd8a88d96c024624", "tree": "8820014555fc76d428b0c15ee3e37f91e5b0729f", "parents": [ "76cd979f4f38a27df22efb5773a0d567181a9392" ], "author": { "name": "Stefan Metzmacher", "email": "metze@samba.org", "time": "Mon Mar 15 12:56:56 2021 +0100" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Mar 18 09:44:35 2021 -0600" }, "message": "io_uring: remove structures from include/linux/io_uring.h\n\nLink: https://lore.kernel.org/r/8c1d14f3748105f4caeda01716d47af2fa41d11c.1615809009.git.metze@samba.org\nSigned-off-by: Stefan Metzmacher \u003cmetze@samba.org\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "76cd979f4f38a27df22efb5773a0d567181a9392", "tree": "3f6e9f655de7ce9756d76ffd9bf459b1bae9ef42", "parents": [ "b7f5a0bfe2061b2c7b2164de06fa4072d7373a45" ], "author": { "name": "Stefan Metzmacher", "email": "metze@samba.org", "time": "Tue Mar 16 16:33:27 2021 +0100" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Mar 18 09:44:06 2021 -0600" }, "message": "io_uring: imply MSG_NOSIGNAL for send[msg]()/recv[msg]() calls\n\nWe never want to generate any SIGPIPE, -EPIPE only is much better.\n\nSigned-off-by: Stefan Metzmacher \u003cmetze@samba.org\u003e\nLink: https://lore.kernel.org/r/38961085c3ec49fd21550c7788f214d1ff02d2d4.1615908477.git.metze@samba.org\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "2c7f76b4c42bd5d953bc821e151644434865f999", "tree": "5133ab0b3059e12e6e4df24c950649bc7d99379c", "parents": [ "0469f2f7ab4c6a6cae4b74c4f981c4da6d909411" ], "author": { "name": "Vitaly Kuznetsov", "email": "vkuznets@redhat.com", "time": "Thu Mar 18 15:09:49 2021 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 18 11:38:58 2021 -0400" }, "message": "selftests: kvm: Add basic Hyper-V clocksources tests\n\nIntroduce a new selftest for Hyper-V clocksources (MSR-based reference TSC\nand TSC page). As a starting point, test the following:\n1) Reference TSC is 1Ghz clock.\n2) Reference TSC and TSC page give the same reading.\n3) TSC page gets updated upon KVM_SET_CLOCK call.\n4) TSC page does not get updated when guest opted for reenlightenment.\n5) Disabled TSC page doesn\u0027t get updated.\n\nSigned-off-by: Vitaly Kuznetsov \u003cvkuznets@redhat.com\u003e\nMessage-Id: \u003c20210318140949.1065740-1-vkuznets@redhat.com\u003e\n[Add a host-side test using TSC + KVM_GET_MSR too. - Paolo]\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "50b1affc891cbc103a2334ce909a026e25f4c84d", "tree": "cc54ff2b8d372402beee2fbd11b18e31fc963f00", "parents": [ "f4df9ee6d7f329d32d756e54f78c084c7e509b24" ], "author": { "name": "Colin Ian King", "email": "colin.king@canonical.com", "time": "Thu Mar 18 13:20:08 2021 +0000" }, "committer": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Thu Mar 18 14:25:13 2021 +0100" }, "message": "ALSA: usb-audio: Fix unintentional sign extension issue\n\nThe shifting of the u8 integer device by 24 bits to the left will\nbe promoted to a 32 bit signed int and then sign-extended to a\n64 bit unsigned long. In the event that the top bit of device is\nset then all then all the upper 32 bits of the unsigned long will\nend up as also being set because of the sign-extension. Fix this\nby casting device to an unsigned long before the shift.\n\nAddresses-Coverity: (\"Unintended sign extension\")\nFixes: a07df82c7990 (\"ALSA: usb-audio: Add DJM750 to Pioneer mixer quirk\")\nSigned-off-by: Colin Ian King \u003ccolin.king@canonical.com\u003e\nLink: https://lore.kernel.org/r/20210318132008.15266-1-colin.king@canonical.com\nSigned-off-by: Takashi Iwai \u003ctiwai@suse.de\u003e\n" }, { "commit": "f4df9ee6d7f329d32d756e54f78c084c7e509b24", "tree": "5a89fe8d7e218989f4bf10eeff1409dc38985ac2", "parents": [ "53b861bec737c189cc14ec3b5785d0f13445ac0f", "f9dc51cc6676e0360dfcb1836b3d426a3ccf3093" ], "author": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Thu Mar 18 14:12:11 2021 +0100" }, "committer": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Thu Mar 18 14:12:11 2021 +0100" }, "message": "Merge tag \u0027asoc-fix-v5.12-rc2\u0027 of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus\n\nASoC: Fixes for v5.12\n\nQuite a lot of mostly platform specific fixes here, the only one which\nis generic is a fix for regressions on devices with more complex\nclocking support with simple-card. There\u0027s also a few new device IDs\nand platform quirks.\n" }, { "commit": "0469f2f7ab4c6a6cae4b74c4f981c4da6d909411", "tree": "b0afc07d99647cf2c4efcbf14438092c0e8e52d0", "parents": [ "cc9cfddb0433961107bb156fa769fdd7eb6718de" ], "author": { "name": "Vitaly Kuznetsov", "email": "vkuznets@redhat.com", "time": "Tue Mar 16 15:37:36 2021 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 18 08:02:46 2021 -0400" }, "message": "KVM: x86: hyper-v: Don\u0027t touch TSC page values when guest opted for re-enlightenment\n\nWhen guest opts for re-enlightenment notifications upon migration, it is\nin its right to assume that TSC page values never change (as they\u0027re only\nsupposed to change upon migration and the host has to keep things as they\nare before it receives confirmation from the guest). This is mostly true\nuntil the guest is migrated somewhere. KVM userspace (e.g. QEMU) will\ntrigger masterclock update by writing to HV_X64_MSR_REFERENCE_TSC, by\ncalling KVM_SET_CLOCK,... and as TSC value and kvmclock reading drift\napart (even slightly), the update causes TSC page values to change.\n\nThe issue at hand is that when Hyper-V is migrated, it uses stale (cached)\nTSC page values to compute the difference between its own clocksource\n(provided by KVM) and its guests\u0027 TSC pages to program synthetic timers\nand in some cases, when TSC page is updated, this puts all stimer\nexpirations in the past. This, in its turn, causes an interrupt storm\nand L2 guests not making much forward progress.\n\nNote, KVM doesn\u0027t fully implement re-enlightenment notification. Basically,\nthe support for reenlightenment MSRs is just a stub and userspace is only\nexpected to expose the feature when TSC scaling on the expected destination\nhosts is available. With TSC scaling, no real re-enlightenment is needed\nas TSC frequency doesn\u0027t change. With TSC scaling becoming ubiquitous, it\nlikely makes little sense to fully implement re-enlightenment in KVM.\n\nPrevent TSC page from being updated after migration. In case it\u0027s not the\nguest who\u0027s initiating the change and when TSC page is already enabled,\njust keep it as it is: TSC value is supposed to be preserved across\nmigration and TSC frequency can\u0027t change with re-enlightenment enabled.\nThe guest is doomed anyway if any of this is not true.\n\nSigned-off-by: Vitaly Kuznetsov \u003cvkuznets@redhat.com\u003e\nMessage-Id: \u003c20210316143736.964151-5-vkuznets@redhat.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "cc9cfddb0433961107bb156fa769fdd7eb6718de", "tree": "94155bc1737da39f522c5b59848fc451c5cb03c1", "parents": [ "e880c6ea55b9805294ecc100ee95e0c9860ae90e" ], "author": { "name": "Vitaly Kuznetsov", "email": "vkuznets@redhat.com", "time": "Tue Mar 16 15:37:35 2021 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 18 08:02:46 2021 -0400" }, "message": "KVM: x86: hyper-v: Track Hyper-V TSC page status\n\nCreate an infrastructure for tracking Hyper-V TSC page status, i.e. if it\nwas updated from guest/host side or if we\u0027ve failed to set it up (because\ne.g. guest wrote some garbage to HV_X64_MSR_REFERENCE_TSC) and there\u0027s no\nneed to retry.\n\nAlso, in a hypothetical situation when we are in \u0027always catchup\u0027 mode for\nTSC we can now avoid contending \u0027hv-\u003ehv_lock\u0027 on every guest enter by\nsetting the state to HV_TSC_PAGE_BROKEN after compute_tsc_page_parameters()\nreturns false.\n\nCheck for HV_TSC_PAGE_SET state instead of \u0027!hv-\u003etsc_ref.tsc_sequence\u0027 in\nget_time_ref_counter() to properly handle the situation when we failed to\nwrite the updated TSC page values to the guest.\n\nSigned-off-by: Vitaly Kuznetsov \u003cvkuznets@redhat.com\u003e\nMessage-Id: \u003c20210316143736.964151-4-vkuznets@redhat.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "8dfd0fa6ecdc5e2099a57d485b7ce237abc6c7a0", "tree": "6ba404a4f44c9a49c63fe13f41f8a152f445698f", "parents": [ "4b8ef157ca832f812b3302b1800548bd92c207de" ], "author": { "name": "Dmitry Osipenko", "email": "digetx@gmail.com", "time": "Fri Mar 12 18:54:39 2021 +0300" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Thu Mar 18 11:31:12 2021 +0100" }, "message": "iommu/tegra-smmu: Make tegra_smmu_probe_device() to handle all IOMMU phandles\n\nThe tegra_smmu_probe_device() handles only the first IOMMU device-tree\nphandle, skipping the rest. Devices like 3D module on Tegra30 have\nmultiple IOMMU phandles, one for each h/w block, and thus, only one\nIOMMU phandle is added to fwspec for the 3D module, breaking GPU.\nPreviously this problem was masked by tegra_smmu_attach_dev() which\ndidn\u0027t use the fwspec, but parsed the DT by itself. The previous commit\nto tegra-smmu driver partially reverted changes that caused problems for\nT124 and now we have tegra_smmu_attach_dev() that uses the fwspec and\nthe old-buggy variant of tegra_smmu_probe_device() which skips secondary\nIOMMUs.\n\nMake tegra_smmu_probe_device() not to skip the secondary IOMMUs. This\nfixes a partially attached IOMMU of the 3D module on Tegra30 and now GPU\nworks properly once again.\n\nFixes: 765a9d1d02b2 (\"iommu/tegra-smmu: Fix mc errors on tegra124-nyan\")\nSigned-off-by: Dmitry Osipenko \u003cdigetx@gmail.com\u003e\nTested-by: Nicolin Chen \u003cnicoleotsuka@gmail.com\u003e\nLink: https://lore.kernel.org/r/20210312155439.18477-1-digetx@gmail.com\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\n" }, { "commit": "4b8ef157ca832f812b3302b1800548bd92c207de", "tree": "d62618dcfc582d352dda0bfb64e4a866d12a8d80", "parents": [ "9f81ca8d1fd68f5697c201f26632ed622e9e462f" ], "author": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Wed Mar 17 10:10:37 2021 +0100" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Thu Mar 18 10:46:49 2021 +0100" }, "message": "iommu/amd: Keep track of amd_iommu_irq_remap state\n\nThe amd_iommu_irq_remap variable is set to true in amd_iommu_prepare().\nBut if initialization fails it is not set to false. Fix that and\ncorrectly keep track of whether irq remapping is enabled or not.\n\nReferences: https://bugzilla.kernel.org/show_bug.cgi?id\u003d212133\nReferences: https://bugzilla.suse.com/show_bug.cgi?id\u003d1183132\nFixes: b34f10c2dc59 (\"iommu/amd: Stop irq_remapping_select() matching when remapping is disabled\")\nCc: stable@vger.kernel.org # v5.11\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\nLink: https://lore.kernel.org/r/20210317091037.31374-4-joro@8bytes.org\nAcked-by: Huang Rui \u003cray.huang@amd.com\u003e\n" }, { "commit": "9f81ca8d1fd68f5697c201f26632ed622e9e462f", "tree": "71274f0650ffe8f11a1ec2adaac3ddfe5235a468", "parents": [ "072a03e0a0b1bc22eb5970727877264657c61fd3" ], "author": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Wed Mar 17 10:10:36 2021 +0100" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Thu Mar 18 10:46:42 2021 +0100" }, "message": "iommu/amd: Don\u0027t call early_amd_iommu_init() when AMD IOMMU is disabled\n\nDon\u0027t even try to initialize the AMD IOMMU hardware when amd_iommu\u003doff has been\npassed on the kernel command line.\n\nReferences: https://bugzilla.kernel.org/show_bug.cgi?id\u003d212133\nReferences: https://bugzilla.suse.com/show_bug.cgi?id\u003d1183132\nCc: stable@vger.kernel.org # v5.11\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\nLink: https://lore.kernel.org/r/20210317091037.31374-3-joro@8bytes.org\nAcked-by: Huang Rui \u003cray.huang@amd.com\u003e\n" }, { "commit": "072a03e0a0b1bc22eb5970727877264657c61fd3", "tree": "44ad742ef357029bb4c35a55b5866d030b1879d4", "parents": [ "1e28eed17697bcf343c6743f0028cc3b5dd88bf0" ], "author": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Wed Mar 17 10:10:35 2021 +0100" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Thu Mar 18 10:46:30 2021 +0100" }, "message": "iommu/amd: Move Stoney Ridge check to detect_ivrs()\n\nThe AMD IOMMU will not be enabled on AMD Stoney Ridge systems. Bail\nout even earlier and refuse to even detect the IOMMU there.\n\nReferences: https://bugzilla.kernel.org/show_bug.cgi?id\u003d212133\nReferences: https://bugzilla.suse.com/show_bug.cgi?id\u003d1183132\nCc: stable@vger.kernel.org # v5.11\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\nLink: https://lore.kernel.org/r/20210317091037.31374-2-joro@8bytes.org\nAcked-by: Huang Rui \u003cray.huang@amd.com\u003e\n" }, { "commit": "2cafd46a714af1e55354bc6dcea9dcc13f9475b5", "tree": "39794cff98ba1a45fb90fa151a09e7f1a3a92ed0", "parents": [ "d2d106fe3badfc3bf0dd3899d1c3f210c7203eab" ], "author": { "name": "Edmundo Carmona Antoranz", "email": "eantoranz@gmail.com", "time": "Tue Mar 16 12:17:35 2021 -0600" }, "committer": { "name": "Greg Kroah-Hartman", "email": "gregkh@linuxfoundation.org", "time": "Thu Mar 18 10:21:46 2021 +0100" }, "message": "staging: vt665x: fix alignment constraints\n\nRemoving 2 instances of alignment warnings\n\ndrivers/staging/vt6655/rxtx.h:153:1: warning: alignment 1 of ‘struct vnt_cts’ is less than 2 [-Wpacked-not-aligned]\ndrivers/staging/vt6655/rxtx.h:163:1: warning: alignment 1 of ‘struct vnt_cts_fb’ is less than 2 [-Wpacked-not-aligned]\n\nThe root cause seems to be that _because_ struct ieee80211_cts is marked as __aligned(2),\nthis requires any encapsulating struct to also have an alignment of 2.\n\nFixes: 2faf12c57efe (\"staging: vt665x: fix alignment constraints\")\nReviewed-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nSigned-off-by: Edmundo Carmona Antoranz \u003ceantoranz@gmail.com\u003e\nLink: https://lore.kernel.org/r/20210316181736.2553318-1-eantoranz@gmail.com\nSigned-off-by: Greg Kroah-Hartman \u003cgregkh@linuxfoundation.org\u003e\n" }, { "commit": "2b8c956ea6ba896ec18ae36c2684ecfa04c1f479", "tree": "66ec4240b95e7144ee0b141d21522f9a8910ac1d", "parents": [ "f09ddcfcb8c569675066337adac2ac205113471f" ], "author": { "name": "Badhri Jagan Sridharan", "email": "badhri@google.com", "time": "Wed Mar 17 23:48:05 2021 -0700" }, "committer": { "name": "Greg Kroah-Hartman", "email": "gregkh@linuxfoundation.org", "time": "Thu Mar 18 09:47:47 2021 +0100" }, "message": "usb: typec: tcpm: Skip sink_cap query only when VDM sm is busy\n\nWhen port partner responds \"Not supported\" to the DiscIdentity command,\nVDM state machine can remain in NVDM_STATE_ERR_TMOUT and this causes\nquerying sink cap to be skipped indefinitely. Hence check for\nvdm_sm_running instead of checking for VDM_STATE_DONE.\n\nFixes: 8dc4bd073663f (\"usb: typec: tcpm: Add support for Sink Fast Role SWAP(FRS)\")\nAcked-by: Heikki Krogerus \u003cheikki.krogerus@linux.intel.com\u003e\nSigned-off-by: Badhri Jagan Sridharan \u003cbadhri@google.com\u003e\nCc: stable \u003cstable@vger.kernel.org\u003e\nLink: https://lore.kernel.org/r/20210318064805.3747831-1-badhri@google.com\nSigned-off-by: Greg Kroah-Hartman \u003cgregkh@linuxfoundation.org\u003e\n" }, { "commit": "bac04454ef9fada009f0572576837548b190bf94", "tree": "d40b2ebf6117bcfe97bad93a2d02d5ec46fda852", "parents": [ "d218a8a3003e84ab136e69a4e30dd4ec7dab2d22" ], "author": { "name": "Elad Grupi", "email": "elad.grupi@dell.com", "time": "Tue Mar 16 17:44:25 2021 +0200" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 18 05:39:18 2021 +0100" }, "message": "nvmet-tcp: fix kmap leak when data digest in use\n\nWhen data digest is enabled we should unmap pdu iovec before handling\nthe data digest pdu.\n\nSigned-off-by: Elad Grupi \u003celad.grupi@dell.com\u003e\nReviewed-by: Sagi Grimberg \u003csagi@grimberg.me\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "d218a8a3003e84ab136e69a4e30dd4ec7dab2d22", "tree": "545ce90ce1854d239f5797a256b22c5a3d04b1f9", "parents": [ "c4c6df5fc84659690d4391d1fba155cd94185295" ], "author": { "name": "Sagi Grimberg", "email": "sagi@grimberg.me", "time": "Mon Mar 15 15:34:51 2021 -0700" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 18 05:39:02 2021 +0100" }, "message": "nvmet: don\u0027t check iosqes,iocqes for discovery controllers\n\nFrom the base spec, Figure 78:\n\n \"Controller Configuration, these fields are defined as parameters to\n configure an \"I/O Controller (IOC)\" and not to configure a \"Discovery\n Controller (DC).\n\n ...\n If the controller does not support I/O queues, then this field shall\n be read-only with a value of 0h\n\nJust perform this check for I/O controllers.\n\nFixes: a07b4970f464 (\"nvmet: add a generic NVMe target\")\nReported-by: Belanger, Martin \u003cMartin.Belanger@dell.com\u003e\nSigned-off-by: Sagi Grimberg \u003csagi@grimberg.me\u003e\nReviewed-by: Chaitanya Kulkarni \u003cchaitanya.kulkarni@wdc.com\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "c4c6df5fc84659690d4391d1fba155cd94185295", "tree": "308001a7f823c5e4822b5ac3292c23a3abddfa1b", "parents": [ "72f572428b83d0bc7028e7c4326d1a5f45205e44" ], "author": { "name": "Sagi Grimberg", "email": "sagi@grimberg.me", "time": "Mon Mar 15 14:04:27 2021 -0700" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 18 05:39:01 2021 +0100" }, "message": "nvme-rdma: fix possible hang when failing to set io queues\n\nWe only setup io queues for nvme controllers, and it makes absolutely no\nsense to allow a controller (re)connect without any I/O queues. If we\nhappen to fail setting the queue count for any reason, we should not allow\nthis to be a successful reconnect as I/O has no chance in going through.\nInstead just fail and schedule another reconnect.\n\nReported-by: Chao Leng \u003clengchao@huawei.com\u003e\nFixes: 711023071960 (\"nvme-rdma: add a NVMe over Fabrics RDMA host driver\")\nSigned-off-by: Sagi Grimberg \u003csagi@grimberg.me\u003e\nReviewed-by: Chao Leng \u003clengchao@huawei.com\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "72f572428b83d0bc7028e7c4326d1a5f45205e44", "tree": "94d24abf670768df1cc10d97974c08598a5948c2", "parents": [ "bb83337058a7000644cdeffc67361d2473534756" ], "author": { "name": "Sagi Grimberg", "email": "sagi@grimberg.me", "time": "Mon Mar 15 14:04:26 2021 -0700" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 18 05:39:01 2021 +0100" }, "message": "nvme-tcp: fix possible hang when failing to set io queues\n\nWe only setup io queues for nvme controllers, and it makes absolutely no\nsense to allow a controller (re)connect without any I/O queues. If we\nhappen to fail setting the queue count for any reason, we should not\nallow this to be a successful reconnect as I/O has no chance in going\nthrough. Instead just fail and schedule another reconnect.\n\nFixes: 3f2304f8c6d6 (\"nvme-tcp: add NVMe over TCP host driver\")\nSigned-off-by: Sagi Grimberg \u003csagi@grimberg.me\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "bb83337058a7000644cdeffc67361d2473534756", "tree": "5632a838e0489afa182e4b366853a60cc5b46c39", "parents": [ "fd0823f405090f9f410fc3e3ff7efb52e7b486fa" ], "author": { "name": "Sagi Grimberg", "email": "sagi@grimberg.me", "time": "Mon Mar 15 13:53:47 2021 -0700" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 18 05:39:01 2021 +0100" }, "message": "nvme-tcp: fix misuse of __smp_processor_id with preemption enabled\n\nFor our pure advisory use-case, we only rely on this call as a hint, so\nfix the warning complaints of using the smp_processor_id variants with\npreemption enabled.\n\nFixes: db5ad6b7f8cd (\"nvme-tcp: try to send request in queue_rq context\")\nFixes: ada831772188 (\"nvme-tcp: Fix warning with CONFIG_DEBUG_PREEMPT\")\nSigned-off-by: Sagi Grimberg \u003csagi@grimberg.me\u003e\nReviewed-by: Chaitanya Kulkarni \u003cchaitanya.kulkarni@wdc.com\u003e\nTested-by: Yi Zhang \u003cyi.zhang@redhat.com\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "fd0823f405090f9f410fc3e3ff7efb52e7b486fa", "tree": "f1449dc8e8b3931ecc45327c6942d7f1c683e2d9", "parents": [ "b94e8cd2e6a94fc7563529ddc82726a7e77e04de" ], "author": { "name": "Sagi Grimberg", "email": "sagi@grimberg.me", "time": "Mon Mar 15 14:08:11 2021 -0700" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 18 05:39:01 2021 +0100" }, "message": "nvme-tcp: fix a NULL deref when receiving a 0-length r2t PDU\n\nWhen the controller sends us a 0-length r2t PDU we should not attempt to\ntry to set up a h2cdata PDU but rather conclude that this is a buggy\ncontroller (forward progress is not possible) and simply fail it\nimmediately.\n\nFixes: 3f2304f8c6d6 (\"nvme-tcp: add NVMe over TCP host driver\")\nReported-by: Belanger, Martin \u003cMartin.Belanger@dell.com\u003e\nSigned-off-by: Sagi Grimberg \u003csagi@grimberg.me\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "b94e8cd2e6a94fc7563529ddc82726a7e77e04de", "tree": "0631a463ddd9ff1b7ace858bc245b5247f1d885f", "parents": [ "985c5a329dfe5ecb782551cddef48912961b83f1" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Mon Mar 15 10:32:07 2021 +0100" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 18 05:38:49 2021 +0100" }, "message": "nvme: fix Write Zeroes limitations\n\nWe voluntarily limit the Write Zeroes sizes to the MDTS value provided by\nthe hardware, but currently get the units wrong, so fix that.\n\nFixes: 6e02318eaea5 (\"nvme: add support for the Write Zeroes command\")\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Keith Busch \u003ckbusch@kernel.org\u003e\nTested-by: Klaus Jensen \u003ck.jensen@samsung.com\u003e\nReviewed-by: Klaus Jensen \u003ck.jensen@samsung.com\u003e\nReviewed-by: Chaitanya Kulkarni \u003cchaitanya.kulkarni@wdc.com\u003e\nReviewed-by: Himanshu Madhani \u003chimanshu.madhani@oracle.com\u003e\n" }, { "commit": "985c5a329dfe5ecb782551cddef48912961b83f1", "tree": "4eb45cb47126dedbb3c4c20d8fcaf651b03aac16", "parents": [ "06c3c3365b4bae5ef0f0525d3683b73cbae1e69c" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Wed Mar 03 13:51:47 2021 +0100" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 18 05:38:48 2021 +0100" }, "message": "nvme: allocate the keep alive request using BLK_MQ_REQ_NOWAIT\n\nTo avoid an error recovery deadlock where the keep alive work is waiting\nfor a request and thus can\u0027t be flushed to make progress for tearing down\nthe controller. Also print the error code returned from\nblk_mq_alloc_request to help debugging any future issues in this code.\n\nBased on an earlier patch from Hannes Reinecke \u003chare@suse.de\u003e.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Sagi Grimberg \u003csagi@grimberg.me\u003e\nReviewed-by: Chaitanya Kulkarni \u003cchaitanya.kulkarni@wdc.com\u003e\nReviewed-by: Hannes Reinecke \u003chare@suse.de\u003e\nReviewed-by: Daniel Wagner \u003cdwagner@suse.de\u003e\n" }, { "commit": "06c3c3365b4bae5ef0f0525d3683b73cbae1e69c", "tree": "528be67c552a37e911476a2cb32f8b042ac57526", "parents": [ "ed01fee283a067c72b2d6500046080dbc1bb9dae" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Wed Mar 03 13:46:06 2021 +0100" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 18 05:38:48 2021 +0100" }, "message": "nvme: merge nvme_keep_alive into nvme_keep_alive_work\n\nMerge nvme_keep_alive into its only caller to prepare for additional\nchanges to this code.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Sagi Grimberg \u003csagi@grimberg.me\u003e\nReviewed-by: Chaitanya Kulkarni \u003cchaitanya.kulkarni@wdc.com\u003e\nReviewed-by: Hannes Reinecke \u003chare@suse.de\u003e\nReviewed-by: Daniel Wagner \u003cdwagner@suse.de\u003e\n" }, { "commit": "ed01fee283a067c72b2d6500046080dbc1bb9dae", "tree": "60b0ad1e023db4800c923331b358a9ef3773d4e0", "parents": [ "1e28eed17697bcf343c6743f0028cc3b5dd88bf0" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Wed Mar 03 13:28:22 2021 +0100" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 18 05:38:48 2021 +0100" }, "message": "nvme-fabrics: only reserve a single tag\n\nFabrics drivers currently reserve two tags on the admin queue. But\ngiven that the connect command is only run on a freshly created queue\nor after all commands have been force aborted we only need to reserve\na single tag.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Sagi Grimberg \u003csagi@grimberg.me\u003e\nReviewed-by: Chaitanya Kulkarni \u003cchaitanya.kulkarni@wdc.com\u003e\nReviewed-by: Hannes Reinecke \u003chare@suse.de\u003e\nReviewed-by: Daniel Wagner \u003cdwagner@suse.de\u003e\n" }, { "commit": "beb6b2f97e0a02164c7f0df6e08c49219cfc2b80", "tree": "3d5b5e2413682ae844ea2e338681fba119599ab2", "parents": [ "34fa493a565cc6fcee6919787c11e264f55603c6" ], "author": { "name": "Dillon Varone", "email": "dillon.varone@amd.com", "time": "Thu Mar 04 17:39:10 2021 -0500" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Thu Mar 18 00:05:22 2021 -0400" }, "message": "drm/amd/display: Remove MPC gamut remap logic for DCN30\n\n[Why?]\nShould only reroute gamut remap to mpc unless 3D LUT is not used and all\nplanes are using the same src-\u003edest.\n\n[How?]\nRemove DCN30 specific logic for rerouting gamut remap to mpc.\n\nBug: https://gitlab.freedesktop.org/drm/amd/-/issues/1513\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Dillon Varone \u003cdillon.varone@amd.com\u003e\nReviewed-by: Krunoslav Kovac \u003cKrunoslav.Kovac@amd.com\u003e\nAcked-by: Aric Cyr \u003cAric.Cyr@amd.com\u003e\nAcked-by: Solomon Chiu \u003csolomon.chiu@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nCc: stable@vger.kernel.org\n" } ], "next": "34fa493a565cc6fcee6919787c11e264f55603c6" }